diff --git "a/checkpoint-534240/trainer_state.json" "b/checkpoint-534240/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-534240/trainer_state.json"
@@ -0,0 +1,6339 @@
+{
+  "best_metric": 3.8620269298553467,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/transformer/1/checkpoints/checkpoint-534240",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 534240,
+  "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": 11.0349,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8119,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.188,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9793,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8074,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6987,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5939,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5293,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4394,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3947,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3436,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3029,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2501,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2025,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1605,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1281,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.0983,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0846,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0412,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.021,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0107,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9653,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9416,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9255,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9185,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8828,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8641,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8569,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8349,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8096,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7946,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740117521192533e-05,
+      "loss": 4.7898,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731731573682013e-05,
+      "loss": 4.7776,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7744,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7505,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7358,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.74,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7022,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.6978,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6838,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6831,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6613,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6745,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6423,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6471,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6362,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6279,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6255,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6015,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 4.5886,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6025,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5884,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5707,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 4.5589,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5672,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5503,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5625,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5352,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5332,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5366,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5176,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5018,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5068,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4904,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4913,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5002,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4929,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4799,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4668,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4594,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4726,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.467,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396441082494794e-05,
+      "loss": 4.4599,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388055134984274e-05,
+      "loss": 4.4575,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4619,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4541,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4283,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4385,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4344,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.42,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4327,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4224,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.429,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.418,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.392,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928747290126923e-05,
+      "loss": 4.4016,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927908695375871e-05,
+      "loss": 4.4018,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927070100624819e-05,
+      "loss": 4.4014,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.394,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3858,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3892,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3773,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9228804026303046e-05,
+      "loss": 4.3709,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922041807879253e-05,
+      "loss": 4.3721,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921203213128201e-05,
+      "loss": 4.368,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.379,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3552,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3514,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3505,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3605,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916173282502262e-05,
+      "loss": 4.3566,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91533468775121e-05,
+      "loss": 4.3432,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914496093000158e-05,
+      "loss": 4.3415,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.3243,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3379,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3236,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3204,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3271,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3251,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3197,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3161,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3088,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3154,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.3027,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.303,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3086,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.3059,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019237232558704e-05,
+      "loss": 4.3034,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010851285048184e-05,
+      "loss": 4.2953,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900246533753767e-05,
+      "loss": 4.2997,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899407939002715e-05,
+      "loss": 4.2948,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.2816,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2985,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2837,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2912,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2744,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2699,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2664,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.27,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2611,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2774,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2738,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2583,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.2609,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2657,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2613,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2666,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2625,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2489,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2541,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.25,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2412,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2518,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2394,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2287,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2524,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2314,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2292,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2312,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.2241,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.197596073150635,
+      "eval_runtime": 303.2397,
+      "eval_samples_per_second": 1258.381,
+      "eval_steps_per_second": 39.325,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2094,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2203,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2365,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2139,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2286,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2034,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.2152,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.1994,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2105,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.206,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.2093,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.207,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.1923,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.1855,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.1848,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.1894,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.1964,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.1837,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1906,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.206,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1814,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.1752,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.1791,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1889,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1685,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.1742,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.1736,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.1675,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.1541,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.161,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1567,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1667,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.1729,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.1561,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.1534,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1656,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844078702117387e-05,
+      "loss": 4.1412,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843240107366335e-05,
+      "loss": 4.1478,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1387,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.145,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.1377,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.1552,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1406,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1414,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.1376,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.1422,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1352,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1264,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834020478745882e-05,
+      "loss": 4.1185,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331818839948295e-05,
+      "loss": 4.1403,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.1304,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.1191,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830667737622047e-05,
+      "loss": 4.1173,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298291428709944e-05,
+      "loss": 4.1257,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289905481199424e-05,
+      "loss": 4.1144,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.1291,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.1141,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.1153,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825637806996108e-05,
+      "loss": 4.1266,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824799212245056e-05,
+      "loss": 4.1049,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.102,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823123660623325e-05,
+      "loss": 4.1116,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822285065872273e-05,
+      "loss": 4.0943,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821446471121221e-05,
+      "loss": 4.1002,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820609514250542e-05,
+      "loss": 4.1124,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.1143,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818932324748438e-05,
+      "loss": 4.1028,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818093729997386e-05,
+      "loss": 4.0966,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.0864,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.1062,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.104,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147409888735514e-05,
+      "loss": 4.1037,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813904032002872e-05,
+      "loss": 4.1042,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81306543725182e-05,
+      "loss": 4.1066,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.1062,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811388247749716e-05,
+      "loss": 4.0819,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810551290879037e-05,
+      "loss": 4.0976,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809712696127985e-05,
+      "loss": 4.0966,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808874101376933e-05,
+      "loss": 4.0811,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.1023,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807198549755203e-05,
+      "loss": 4.0945,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806359955004151e-05,
+      "loss": 4.1003,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805521360253099e-05,
+      "loss": 4.0971,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804682765502047e-05,
+      "loss": 4.0739,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8038458086313677e-05,
+      "loss": 4.0847,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030072138803156e-05,
+      "loss": 4.0838,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021686191292636e-05,
+      "loss": 4.0881,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013300243782116e-05,
+      "loss": 4.0832,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004914296271596e-05,
+      "loss": 4.0752,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996544727564805e-05,
+      "loss": 4.0851,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988158780054285e-05,
+      "loss": 4.0794,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979789211347494e-05,
+      "loss": 4.0653,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797140326383698e-05,
+      "loss": 4.0728,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796301731632646e-05,
+      "loss": 4.0785,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795463136881594e-05,
+      "loss": 4.0875,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794624542130542e-05,
+      "loss": 4.0694,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79378594737949e-05,
+      "loss": 4.0568,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792947352628438e-05,
+      "loss": 4.0675,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792108757877386e-05,
+      "loss": 4.0752,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79127343888708e-05,
+      "loss": 4.0757,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790434844136028e-05,
+      "loss": 4.0686,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789596249384976e-05,
+      "loss": 4.0625,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788757654633924e-05,
+      "loss": 4.0524,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787919059882872e-05,
+      "loss": 4.0657,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708046513182e-05,
+      "loss": 4.0519,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786241870380768e-05,
+      "loss": 4.0563,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854049135100895e-05,
+      "loss": 4.0615,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845663187590375e-05,
+      "loss": 4.0592,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837277240079855e-05,
+      "loss": 4.055,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828891292569335e-05,
+      "loss": 4.0609,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820505345058815e-05,
+      "loss": 4.0503,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812119397548295e-05,
+      "loss": 4.0559,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803733450037775e-05,
+      "loss": 4.0504,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779534750252725e-05,
+      "loss": 4.0482,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786977933820464e-05,
+      "loss": 4.0606,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778591986309944e-05,
+      "loss": 4.0592,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0548,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0476,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0524,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450645750716e-05,
+      "loss": 4.0508,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773667862756108e-05,
+      "loss": 4.0398,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772829268005055e-05,
+      "loss": 4.0598,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771990673254003e-05,
+      "loss": 4.0436,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771152078502951e-05,
+      "loss": 4.0526,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770315121632272e-05,
+      "loss": 4.0374,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76947652688122e-05,
+      "loss": 4.0377,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768637932130168e-05,
+      "loss": 4.0333,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767799337379116e-05,
+      "loss": 4.035,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766962380508437e-05,
+      "loss": 4.0335,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766123785757386e-05,
+      "loss": 4.0478,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765285191006334e-05,
+      "loss": 4.0488,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764446596255282e-05,
+      "loss": 4.0361,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636096393846026e-05,
+      "loss": 4.0292,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7627710446335506e-05,
+      "loss": 4.0414,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7619324498824986e-05,
+      "loss": 4.0397,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610938551314466e-05,
+      "loss": 4.0463,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602568982607675e-05,
+      "loss": 4.04,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594183035097155e-05,
+      "loss": 4.0343,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585797087586635e-05,
+      "loss": 4.029,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577411140076115e-05,
+      "loss": 4.0395,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569041571369324e-05,
+      "loss": 4.0283,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756065562385881e-05,
+      "loss": 4.0324,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755226967634829e-05,
+      "loss": 4.0301,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754388372883777e-05,
+      "loss": 4.0212,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753551416013098e-05,
+      "loss": 4.0435,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752712821262046e-05,
+      "loss": 4.0235,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751874226510994e-05,
+      "loss": 4.0202,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751035631759942e-05,
+      "loss": 4.0227,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750198674889263e-05,
+      "loss": 4.0185,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.028837203979492,
+      "eval_runtime": 303.893,
+      "eval_samples_per_second": 1255.675,
+      "eval_steps_per_second": 39.241,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749360080138211e-05,
+      "loss": 4.0037,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748521485387159e-05,
+      "loss": 4.0188,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747682890636107e-05,
+      "loss": 4.0332,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746844295885055e-05,
+      "loss": 4.0175,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746005701134003e-05,
+      "loss": 4.0303,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745167106382951e-05,
+      "loss": 4.0088,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744328511631899e-05,
+      "loss": 4.0208,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434899168808475e-05,
+      "loss": 4.0015,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.0181,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.0142,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409757705080644e-05,
+      "loss": 4.0141,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401371757570124e-05,
+      "loss": 4.0229,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739300218886333e-05,
+      "loss": 4.0026,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738461624135281e-05,
+      "loss": 3.9977,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737623029384229e-05,
+      "loss": 3.9976,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736784434633177e-05,
+      "loss": 4.0064,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735945839882125e-05,
+      "loss": 4.0094,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735107245131073e-05,
+      "loss": 4.0025,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734268650380021e-05,
+      "loss": 4.0066,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733430055628969e-05,
+      "loss": 4.0246,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732591460877917e-05,
+      "loss": 3.9993,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731752866126865e-05,
+      "loss": 3.9954,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730914271375813e-05,
+      "loss": 4.0023,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730077314505134e-05,
+      "loss": 4.0084,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729238719754082e-05,
+      "loss": 3.9934,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728401762883403e-05,
+      "loss": 3.9974,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727563168132351e-05,
+      "loss": 3.9979,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726724573381299e-05,
+      "loss": 3.9935,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725885978630247e-05,
+      "loss": 3.9826,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725049021759568e-05,
+      "loss": 3.9901,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724210427008517e-05,
+      "loss": 3.989,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233718322574647e-05,
+      "loss": 3.9929,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225332375064127e-05,
+      "loss": 4.0066,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216946427553606e-05,
+      "loss": 3.9853,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208560480043086e-05,
+      "loss": 3.9856,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200174532532566e-05,
+      "loss": 3.9973,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191788585022046e-05,
+      "loss": 3.9807,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183419016315255e-05,
+      "loss": 3.9787,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175033068804735e-05,
+      "loss": 3.9743,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166647121294215e-05,
+      "loss": 3.9843,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158261173783695e-05,
+      "loss": 3.9732,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149891605076904e-05,
+      "loss": 3.9965,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141505657566384e-05,
+      "loss": 3.9753,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133119710055864e-05,
+      "loss": 3.9859,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712475014134908e-05,
+      "loss": 3.9805,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711636419383856e-05,
+      "loss": 3.9802,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710797824632804e-05,
+      "loss": 3.9787,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709959229881752e-05,
+      "loss": 3.9669,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709122273011073e-05,
+      "loss": 3.9646,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708283678260021e-05,
+      "loss": 3.9848,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707445083508969e-05,
+      "loss": 3.9775,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706606488757917e-05,
+      "loss": 3.9651,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705769531887238e-05,
+      "loss": 3.9684,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704932575016559e-05,
+      "loss": 3.9764,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040939802655074e-05,
+      "loss": 3.9563,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032553855144554e-05,
+      "loss": 3.9761,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024167907634034e-05,
+      "loss": 3.9664,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015781960123514e-05,
+      "loss": 3.9647,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007396012612994e-05,
+      "loss": 3.9826,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6999010065102474e-05,
+      "loss": 3.9578,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990624117591954e-05,
+      "loss": 3.9547,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698225454888516e-05,
+      "loss": 3.9657,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697386860137464e-05,
+      "loss": 3.9502,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696548265386412e-05,
+      "loss": 3.955,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69570967063536e-05,
+      "loss": 3.9678,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694872713764681e-05,
+      "loss": 3.9698,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694034119013629e-05,
+      "loss": 3.9539,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693195524262577e-05,
+      "loss": 3.9561,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692358567391899e-05,
+      "loss": 3.9435,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.9648,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690681377889795e-05,
+      "loss": 3.96,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689842783138743e-05,
+      "loss": 3.9665,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689004188387691e-05,
+      "loss": 3.9611,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688165593636639e-05,
+      "loss": 3.9647,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873286367659596e-05,
+      "loss": 3.9722,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864900420149076e-05,
+      "loss": 3.9437,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856514472638556e-05,
+      "loss": 3.9585,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848128525128036e-05,
+      "loss": 3.9628,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683974257761751e-05,
+      "loss": 3.9447,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831356630106996e-05,
+      "loss": 3.965,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822970682596476e-05,
+      "loss": 3.9592,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814584735085956e-05,
+      "loss": 3.9648,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68062315451829e-05,
+      "loss": 3.9625,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679784559767238e-05,
+      "loss": 3.9401,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678945965016186e-05,
+      "loss": 3.9525,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678107370265134e-05,
+      "loss": 3.9503,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772687755140814e-05,
+      "loss": 3.9584,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764301807630294e-05,
+      "loss": 3.9487,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675593223892351e-05,
+      "loss": 3.9471,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674754629141298e-05,
+      "loss": 3.9532,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673916034390246e-05,
+      "loss": 3.95,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673077439639194e-05,
+      "loss": 3.936,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672238844888143e-05,
+      "loss": 3.9429,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671400250137091e-05,
+      "loss": 3.9526,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670563293266412e-05,
+      "loss": 3.9567,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66972469851536e-05,
+      "loss": 3.9462,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668886103764308e-05,
+      "loss": 3.9278,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668047509013256e-05,
+      "loss": 3.9411,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667210552142577e-05,
+      "loss": 3.948,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666371957391525e-05,
+      "loss": 3.9476,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665533362640473e-05,
+      "loss": 3.9441,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664694767889421e-05,
+      "loss": 3.9383,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663856173138369e-05,
+      "loss": 3.9274,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663017578387317e-05,
+      "loss": 3.9463,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662178983636265e-05,
+      "loss": 3.9233,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661340388885213e-05,
+      "loss": 3.9363,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660503432014534e-05,
+      "loss": 3.9313,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659664837263482e-05,
+      "loss": 3.9434,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65882624251243e-05,
+      "loss": 3.928,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657987647761378e-05,
+      "loss": 3.9404,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657149053010326e-05,
+      "loss": 3.9312,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656310458259274e-05,
+      "loss": 3.9305,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655473501388595e-05,
+      "loss": 3.9359,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654634906637543e-05,
+      "loss": 3.9262,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653796311886491e-05,
+      "loss": 3.941,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652957717135439e-05,
+      "loss": 3.9399,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65212076026476e-05,
+      "loss": 3.9373,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651282165513708e-05,
+      "loss": 3.9297,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650443570762657e-05,
+      "loss": 3.9329,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649604976011605e-05,
+      "loss": 3.9367,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648768019140926e-05,
+      "loss": 3.9187,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647929424389874e-05,
+      "loss": 3.9408,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647090829638822e-05,
+      "loss": 3.9298,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64625223488777e-05,
+      "loss": 3.9347,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645413640136718e-05,
+      "loss": 3.9242,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644575045385665e-05,
+      "loss": 3.9268,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643736450634613e-05,
+      "loss": 3.9145,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428994937639346e-05,
+      "loss": 3.9213,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642060899012882e-05,
+      "loss": 3.9201,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412223042618306e-05,
+      "loss": 3.9341,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403837095107786e-05,
+      "loss": 3.9352,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395467526401e-05,
+      "loss": 3.9255,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387081578890475e-05,
+      "loss": 3.9173,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378695631379955e-05,
+      "loss": 3.9272,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370309683869435e-05,
+      "loss": 3.9288,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636194011516265e-05,
+      "loss": 3.9334,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353554167652124e-05,
+      "loss": 3.9279,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345168220141604e-05,
+      "loss": 3.9191,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336782272631084e-05,
+      "loss": 3.9253,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328396325120564e-05,
+      "loss": 3.9297,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320010377610043e-05,
+      "loss": 3.9167,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311624430099523e-05,
+      "loss": 3.9222,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630325486139274e-05,
+      "loss": 3.9206,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629486891388222e-05,
+      "loss": 3.9121,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62864829663717e-05,
+      "loss": 3.9349,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627809701886118e-05,
+      "loss": 3.9176,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626971107135066e-05,
+      "loss": 3.912,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626134150264387e-05,
+      "loss": 3.9133,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625295555513335e-05,
+      "loss": 3.9098,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9549131393432617,
+      "eval_runtime": 303.0366,
+      "eval_samples_per_second": 1259.224,
+      "eval_steps_per_second": 39.352,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624456960762283e-05,
+      "loss": 3.9038,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623618366011231e-05,
+      "loss": 3.9111,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622781409140552e-05,
+      "loss": 3.9277,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219428143895e-05,
+      "loss": 3.9124,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621104219638448e-05,
+      "loss": 3.9321,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620265624887396e-05,
+      "loss": 3.8982,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619428668016717e-05,
+      "loss": 3.9187,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618590073265665e-05,
+      "loss": 3.8965,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617751478514613e-05,
+      "loss": 3.914,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616912883763561e-05,
+      "loss": 3.9126,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616074289012509e-05,
+      "loss": 3.908,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61523733214183e-05,
+      "loss": 3.9185,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614400375271151e-05,
+      "loss": 3.9001,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613561780520099e-05,
+      "loss": 3.8977,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612723185769047e-05,
+      "loss": 3.8967,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611884591017995e-05,
+      "loss": 3.9037,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611045996266943e-05,
+      "loss": 3.9073,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610207401515891e-05,
+      "loss": 3.9041,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609368806764839e-05,
+      "loss": 3.905,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608530212013788e-05,
+      "loss": 3.9227,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607691617262736e-05,
+      "loss": 3.9021,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068546603920566e-05,
+      "loss": 3.9007,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060160656410046e-05,
+      "loss": 3.9044,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051774708899526e-05,
+      "loss": 3.9083,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043388761389006e-05,
+      "loss": 3.8937,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603500281387848e-05,
+      "loss": 3.9019,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026633245171695e-05,
+      "loss": 3.902,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018247297661175e-05,
+      "loss": 3.893,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009861350150655e-05,
+      "loss": 3.8882,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600147540264013e-05,
+      "loss": 3.8876,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993089455129615e-05,
+      "loss": 3.8973,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984703507619095e-05,
+      "loss": 3.895,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597633393891231e-05,
+      "loss": 3.9109,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967947991401784e-05,
+      "loss": 3.8976,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959562043891264e-05,
+      "loss": 3.8839,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951176096380744e-05,
+      "loss": 3.8999,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594280652767395e-05,
+      "loss": 3.8924,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593442058016343e-05,
+      "loss": 3.8819,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592603463265291e-05,
+      "loss": 3.8831,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591764868514239e-05,
+      "loss": 3.8874,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590926273763187e-05,
+      "loss": 3.882,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590087679012135e-05,
+      "loss": 3.9032,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589250722141457e-05,
+      "loss": 3.8856,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588412127390405e-05,
+      "loss": 3.8918,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587573532639353e-05,
+      "loss": 3.8904,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586734937888301e-05,
+      "loss": 3.8857,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585897981017622e-05,
+      "loss": 3.889,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58505938626657e-05,
+      "loss": 3.8741,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584220791515518e-05,
+      "loss": 3.8659,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583382196764466e-05,
+      "loss": 3.8969,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582543602013414e-05,
+      "loss": 3.8843,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581706645142735e-05,
+      "loss": 3.8796,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580868050391683e-05,
+      "loss": 3.8742,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580029455640631e-05,
+      "loss": 3.882,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579190860889579e-05,
+      "loss": 3.8682,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783539040189e-05,
+      "loss": 3.8846,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577515309267848e-05,
+      "loss": 3.8821,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576676714516796e-05,
+      "loss": 3.8763,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575838119765744e-05,
+      "loss": 3.8896,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575001162895065e-05,
+      "loss": 3.8681,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574162568144013e-05,
+      "loss": 3.8659,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573323973392961e-05,
+      "loss": 3.8812,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572485378641909e-05,
+      "loss": 3.8596,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57164842177123e-05,
+      "loss": 3.8626,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570809827020178e-05,
+      "loss": 3.8827,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569971232269126e-05,
+      "loss": 3.8838,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569132637518074e-05,
+      "loss": 3.8637,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568294042767022e-05,
+      "loss": 3.8755,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674570858963436e-05,
+      "loss": 3.8561,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666184911452916e-05,
+      "loss": 3.8781,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657798963942396e-05,
+      "loss": 3.8734,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649413016431876e-05,
+      "loss": 3.88,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641043447725085e-05,
+      "loss": 3.8726,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632657500214565e-05,
+      "loss": 3.8804,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624271552704045e-05,
+      "loss": 3.8873,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615885605193525e-05,
+      "loss": 3.8571,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607516036486734e-05,
+      "loss": 3.8762,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599130088976214e-05,
+      "loss": 3.879,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590744141465694e-05,
+      "loss": 3.8545,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582358193955174e-05,
+      "loss": 3.8843,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557398862524839e-05,
+      "loss": 3.8759,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556560267773787e-05,
+      "loss": 3.8778,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555721673022735e-05,
+      "loss": 3.8823,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554883078271683e-05,
+      "loss": 3.859,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554044483520631e-05,
+      "loss": 3.8631,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553207526649952e-05,
+      "loss": 3.8684,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523689318989e-05,
+      "loss": 3.8754,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551530337147848e-05,
+      "loss": 3.8665,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550691742396796e-05,
+      "loss": 3.8662,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549854785526117e-05,
+      "loss": 3.8654,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549016190775065e-05,
+      "loss": 3.8745,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548177596024013e-05,
+      "loss": 3.8541,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547339001272961e-05,
+      "loss": 3.8602,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546500406521909e-05,
+      "loss": 3.8709,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54566344965123e-05,
+      "loss": 3.8761,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544824854900178e-05,
+      "loss": 3.8654,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543986260149126e-05,
+      "loss": 3.8482,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431476653980736e-05,
+      "loss": 3.8594,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542310708527395e-05,
+      "loss": 3.8638,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541472113776343e-05,
+      "loss": 3.8673,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540633519025291e-05,
+      "loss": 3.8661,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397949242742385e-05,
+      "loss": 3.8596,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53895796740356e-05,
+      "loss": 3.8491,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538119372652508e-05,
+      "loss": 3.8652,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537280777901456e-05,
+      "loss": 3.8467,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536442183150404e-05,
+      "loss": 3.8566,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535605226279726e-05,
+      "loss": 3.8494,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534766631528674e-05,
+      "loss": 3.8644,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533928036777621e-05,
+      "loss": 3.8504,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533089442026569e-05,
+      "loss": 3.8678,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322524851558906e-05,
+      "loss": 3.8483,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314138904048386e-05,
+      "loss": 3.8537,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530575295653786e-05,
+      "loss": 3.856,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529736700902734e-05,
+      "loss": 3.8525,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528898106151682e-05,
+      "loss": 3.8604,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280611492810035e-05,
+      "loss": 3.8671,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527224192410325e-05,
+      "loss": 3.858,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526385597659273e-05,
+      "loss": 3.8517,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525547002908221e-05,
+      "loss": 3.8573,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247084081571684e-05,
+      "loss": 3.8589,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238698134061164e-05,
+      "loss": 3.8424,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230312186550644e-05,
+      "loss": 3.8634,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221926239040124e-05,
+      "loss": 3.8486,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213540291529604e-05,
+      "loss": 3.8612,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5205154344019084e-05,
+      "loss": 3.8494,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519678477531229e-05,
+      "loss": 3.8513,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518839882780177e-05,
+      "loss": 3.8362,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518001288029125e-05,
+      "loss": 3.8519,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517162693278074e-05,
+      "loss": 3.8445,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516325736407395e-05,
+      "loss": 3.8552,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515487141656343e-05,
+      "loss": 3.8581,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514648546905291e-05,
+      "loss": 3.8535,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513809952154239e-05,
+      "loss": 3.8405,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51297299528356e-05,
+      "loss": 3.8549,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512134400532508e-05,
+      "loss": 3.856,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511295805781456e-05,
+      "loss": 3.8568,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510457211030404e-05,
+      "loss": 3.8543,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096202541597246e-05,
+      "loss": 3.8474,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087816594086726e-05,
+      "loss": 3.8483,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079430646576206e-05,
+      "loss": 3.8577,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507104469906569e-05,
+      "loss": 3.8427,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506265875155517e-05,
+      "loss": 3.8509,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505428918284838e-05,
+      "loss": 3.8433,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504590323533786e-05,
+      "loss": 3.8456,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503751728782734e-05,
+      "loss": 3.8604,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502913134031682e-05,
+      "loss": 3.8473,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502076177161003e-05,
+      "loss": 3.8384,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501237582409951e-05,
+      "loss": 3.8384,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500398987658899e-05,
+      "loss": 3.8424,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9138612747192383,
+      "eval_runtime": 304.2723,
+      "eval_samples_per_second": 1254.11,
+      "eval_steps_per_second": 39.192,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499560392907847e-05,
+      "loss": 3.8267,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498721798156795e-05,
+      "loss": 3.8412,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497883203405743e-05,
+      "loss": 3.8532,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497044608654691e-05,
+      "loss": 3.8441,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496206013903639e-05,
+      "loss": 3.8601,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495367419152587e-05,
+      "loss": 3.8365,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494528824401535e-05,
+      "loss": 3.8433,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493690229650483e-05,
+      "loss": 3.8312,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928532727798047e-05,
+      "loss": 3.8381,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492014678028752e-05,
+      "loss": 3.8429,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911760832777e-05,
+      "loss": 3.8404,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490337488526648e-05,
+      "loss": 3.8489,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895021695363425e-05,
+      "loss": 3.8301,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886635747852905e-05,
+      "loss": 3.8295,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878249800342385e-05,
+      "loss": 3.8281,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869863852831864e-05,
+      "loss": 3.8316,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861477905321344e-05,
+      "loss": 3.8398,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853091957810824e-05,
+      "loss": 3.8321,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844706010300304e-05,
+      "loss": 3.838,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836320062789784e-05,
+      "loss": 3.8535,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827950494082993e-05,
+      "loss": 3.838,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481956454657247e-05,
+      "loss": 3.8333,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481117859906195e-05,
+      "loss": 3.8329,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480279265155143e-05,
+      "loss": 3.8374,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479442308284464e-05,
+      "loss": 3.8255,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478603713533412e-05,
+      "loss": 3.8359,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47776511878236e-05,
+      "loss": 3.8327,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476926524031308e-05,
+      "loss": 3.828,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47608956716063e-05,
+      "loss": 3.8234,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475250972409578e-05,
+      "loss": 3.823,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474412377658526e-05,
+      "loss": 3.8275,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473573782907474e-05,
+      "loss": 3.8302,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472736826036795e-05,
+      "loss": 3.8433,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471898231285743e-05,
+      "loss": 3.8308,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471059636534691e-05,
+      "loss": 3.8184,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470221041783639e-05,
+      "loss": 3.8323,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693840849129596e-05,
+      "loss": 3.8327,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685454901619076e-05,
+      "loss": 3.8125,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677068954108556e-05,
+      "loss": 3.8142,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668683006598036e-05,
+      "loss": 3.8248,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466031343789125e-05,
+      "loss": 3.8155,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465192749038073e-05,
+      "loss": 3.8379,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464354154287021e-05,
+      "loss": 3.8211,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463515559535969e-05,
+      "loss": 3.823,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46267860266529e-05,
+      "loss": 3.8248,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461840007914238e-05,
+      "loss": 3.8259,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461001413163186e-05,
+      "loss": 3.8239,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460162818412134e-05,
+      "loss": 3.8099,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459325861541455e-05,
+      "loss": 3.7994,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458487266790403e-05,
+      "loss": 3.836,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457648672039351e-05,
+      "loss": 3.8172,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456810077288299e-05,
+      "loss": 3.8185,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559731204176205e-05,
+      "loss": 3.8151,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551345256665685e-05,
+      "loss": 3.8129,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542959309155165e-05,
+      "loss": 3.8081,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534573361644645e-05,
+      "loss": 3.8175,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526203792937854e-05,
+      "loss": 3.8194,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517817845427334e-05,
+      "loss": 3.8125,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509431897916814e-05,
+      "loss": 3.831,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501045950406294e-05,
+      "loss": 3.8031,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44926763816995e-05,
+      "loss": 3.8011,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448429043418898e-05,
+      "loss": 3.8238,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447590448667846e-05,
+      "loss": 3.7943,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446751853916794e-05,
+      "loss": 3.7999,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445914897046116e-05,
+      "loss": 3.8227,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445076302295064e-05,
+      "loss": 3.8234,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444237707544012e-05,
+      "loss": 3.802,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44339911279296e-05,
+      "loss": 3.8128,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442562155922281e-05,
+      "loss": 3.7949,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441723561171229e-05,
+      "loss": 3.8142,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440884966420177e-05,
+      "loss": 3.8093,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440046371669125e-05,
+      "loss": 3.8179,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439209414798446e-05,
+      "loss": 3.8104,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438370820047394e-05,
+      "loss": 3.8196,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437532225296342e-05,
+      "loss": 3.8264,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669363054529e-05,
+      "loss": 3.8007,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435856673674611e-05,
+      "loss": 3.8108,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435018078923559e-05,
+      "loss": 3.8195,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434179484172507e-05,
+      "loss": 3.7947,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433340889421455e-05,
+      "loss": 3.8213,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432503932550776e-05,
+      "loss": 3.8197,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431665337799724e-05,
+      "loss": 3.819,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430826743048672e-05,
+      "loss": 3.8208,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42998814829762e-05,
+      "loss": 3.8011,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429151191426941e-05,
+      "loss": 3.7996,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428312596675889e-05,
+      "loss": 3.808,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427474001924837e-05,
+      "loss": 3.8185,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426635407173785e-05,
+      "loss": 3.807,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257984503031066e-05,
+      "loss": 3.8049,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249598555520546e-05,
+      "loss": 3.8044,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241212608010026e-05,
+      "loss": 3.8156,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232826660499506e-05,
+      "loss": 3.7997,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224457091792715e-05,
+      "loss": 3.8,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216071144282195e-05,
+      "loss": 3.8078,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207685196771675e-05,
+      "loss": 3.8193,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199299249261155e-05,
+      "loss": 3.8088,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190929680554364e-05,
+      "loss": 3.7935,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182543733043844e-05,
+      "loss": 3.7993,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174157785533324e-05,
+      "loss": 3.8051,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165771838022804e-05,
+      "loss": 3.8114,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157385890512284e-05,
+      "loss": 3.8061,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41490163218055e-05,
+      "loss": 3.8037,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414063037429498e-05,
+      "loss": 3.7914,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413224442678446e-05,
+      "loss": 3.8026,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412385847927394e-05,
+      "loss": 3.7907,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411548891056715e-05,
+      "loss": 3.8024,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410710296305663e-05,
+      "loss": 3.7853,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409871701554611e-05,
+      "loss": 3.8146,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409034744683932e-05,
+      "loss": 3.7939,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40819614993288e-05,
+      "loss": 3.8024,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407357555181828e-05,
+      "loss": 3.7968,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406518960430776e-05,
+      "loss": 3.7947,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056820035600974e-05,
+      "loss": 3.7992,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048434088090454e-05,
+      "loss": 3.7973,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4040048140579934e-05,
+      "loss": 3.8023,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031662193069414e-05,
+      "loss": 3.8095,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402330900316635e-05,
+      "loss": 3.8007,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401492305565583e-05,
+      "loss": 3.7973,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400653710814531e-05,
+      "loss": 3.8,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399815116063479e-05,
+      "loss": 3.8057,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398976521312427e-05,
+      "loss": 3.7826,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398137926561375e-05,
+      "loss": 3.8063,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397299331810323e-05,
+      "loss": 3.7911,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396460737059271e-05,
+      "loss": 3.8043,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395623780188593e-05,
+      "loss": 3.7941,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394785185437541e-05,
+      "loss": 3.7978,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393946590686489e-05,
+      "loss": 3.7821,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393107995935437e-05,
+      "loss": 3.793,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922710390647576e-05,
+      "loss": 3.7933,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914324443137056e-05,
+      "loss": 3.7972,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905938495626536e-05,
+      "loss": 3.7978,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897552548116016e-05,
+      "loss": 3.8024,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889182979409225e-05,
+      "loss": 3.7832,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880797031898705e-05,
+      "loss": 3.7982,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872411084388185e-05,
+      "loss": 3.8031,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3864025136877665e-05,
+      "loss": 3.8006,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385565556817088e-05,
+      "loss": 3.7958,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384726962066036e-05,
+      "loss": 3.797,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383888367314984e-05,
+      "loss": 3.7902,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830497725639314e-05,
+      "loss": 3.8041,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382212815693253e-05,
+      "loss": 3.7865,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381374220942201e-05,
+      "loss": 3.798,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380535626191148e-05,
+      "loss": 3.7899,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379697031440096e-05,
+      "loss": 3.7907,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378860074569418e-05,
+      "loss": 3.8064,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378021479818366e-05,
+      "loss": 3.7936,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377182885067313e-05,
+      "loss": 3.7887,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376344290316262e-05,
+      "loss": 3.7794,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3755073334455835e-05,
+      "loss": 3.7897,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8884811401367188,
+      "eval_runtime": 302.7215,
+      "eval_samples_per_second": 1260.535,
+      "eval_steps_per_second": 39.393,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3746687386945315e-05,
+      "loss": 3.783,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373830143943479e-05,
+      "loss": 3.7837,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372991549192427e-05,
+      "loss": 3.8004,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721545923217484e-05,
+      "loss": 3.7937,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371315997570696e-05,
+      "loss": 3.8071,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370477402819644e-05,
+      "loss": 3.7853,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369638808068592e-05,
+      "loss": 3.7841,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368801851197913e-05,
+      "loss": 3.782,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3679632564468606e-05,
+      "loss": 3.783,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3671246616958086e-05,
+      "loss": 3.7908,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366286066944757e-05,
+      "loss": 3.7901,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365449110074079e-05,
+      "loss": 3.7922,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3646121532034e-05,
+      "loss": 3.7831,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363773558452348e-05,
+      "loss": 3.7776,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362934963701296e-05,
+      "loss": 3.7759,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362096368950243e-05,
+      "loss": 3.7783,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361257774199191e-05,
+      "loss": 3.7863,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360419179448139e-05,
+      "loss": 3.7835,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359580584697087e-05,
+      "loss": 3.7864,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358741989946035e-05,
+      "loss": 3.8006,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357905033075356e-05,
+      "loss": 3.7885,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357066438324304e-05,
+      "loss": 3.7776,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3562278435732526e-05,
+      "loss": 3.7836,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3553892488222006e-05,
+      "loss": 3.7846,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545522919515215e-05,
+      "loss": 3.7766,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3537136972004695e-05,
+      "loss": 3.7853,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528751024494175e-05,
+      "loss": 3.7809,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520365076983655e-05,
+      "loss": 3.7792,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511995508276864e-05,
+      "loss": 3.7751,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503609560766344e-05,
+      "loss": 3.7671,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3495223613255824e-05,
+      "loss": 3.7771,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486837665745304e-05,
+      "loss": 3.7792,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347846809703851e-05,
+      "loss": 3.7886,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347008214952799e-05,
+      "loss": 3.7814,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346169620201748e-05,
+      "loss": 3.768,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345332663331069e-05,
+      "loss": 3.786,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344494068580017e-05,
+      "loss": 3.7811,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343655473828965e-05,
+      "loss": 3.7598,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342816879077913e-05,
+      "loss": 3.7719,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341979922207234e-05,
+      "loss": 3.7712,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341141327456182e-05,
+      "loss": 3.7688,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34030273270513e-05,
+      "loss": 3.787,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339464137954078e-05,
+      "loss": 3.7727,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338627181083399e-05,
+      "loss": 3.7735,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337788586332347e-05,
+      "loss": 3.7734,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336949991581295e-05,
+      "loss": 3.7747,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3361113968302434e-05,
+      "loss": 3.779,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335274439959564e-05,
+      "loss": 3.7601,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334435845208512e-05,
+      "loss": 3.7461,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33359725045746e-05,
+      "loss": 3.7891,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332758655706408e-05,
+      "loss": 3.768,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331921698835729e-05,
+      "loss": 3.7773,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331083104084677e-05,
+      "loss": 3.7615,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330244509333625e-05,
+      "loss": 3.7647,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329405914582573e-05,
+      "loss": 3.7619,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328568957711894e-05,
+      "loss": 3.7661,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327730362960842e-05,
+      "loss": 3.7723,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32689176820979e-05,
+      "loss": 3.7613,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326053173458739e-05,
+      "loss": 3.7831,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3252162165880596e-05,
+      "loss": 3.7583,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243776218370076e-05,
+      "loss": 3.7519,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3235390270859556e-05,
+      "loss": 3.7755,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3227004323349036e-05,
+      "loss": 3.7508,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218634754642245e-05,
+      "loss": 3.7497,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3210248807131725e-05,
+      "loss": 3.773,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201862859621205e-05,
+      "loss": 3.7706,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193476912110685e-05,
+      "loss": 3.7586,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185107343403894e-05,
+      "loss": 3.7666,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176721395893374e-05,
+      "loss": 3.7466,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168335448382854e-05,
+      "loss": 3.7682,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315994950087234e-05,
+      "loss": 3.7602,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315157993216555e-05,
+      "loss": 3.7725,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314319398465503e-05,
+      "loss": 3.7649,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313480803714451e-05,
+      "loss": 3.7688,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312642208963399e-05,
+      "loss": 3.7802,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31180525209272e-05,
+      "loss": 3.7529,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310966657341668e-05,
+      "loss": 3.7673,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310128062590616e-05,
+      "loss": 3.7685,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309289467839564e-05,
+      "loss": 3.7501,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308452510968885e-05,
+      "loss": 3.7727,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307613916217833e-05,
+      "loss": 3.7756,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306775321466781e-05,
+      "loss": 3.7683,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305936726715729e-05,
+      "loss": 3.7802,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050997698450504e-05,
+      "loss": 3.7534,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042611750939984e-05,
+      "loss": 3.7509,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034225803429464e-05,
+      "loss": 3.7632,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025839855918944e-05,
+      "loss": 3.7665,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301747028721215e-05,
+      "loss": 3.7668,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300908433970163e-05,
+      "loss": 3.7577,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300069839219111e-05,
+      "loss": 3.755,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299231244468059e-05,
+      "loss": 3.7691,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298395925477753e-05,
+      "loss": 3.754,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297557330726702e-05,
+      "loss": 3.7541,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29671873597565e-05,
+      "loss": 3.7618,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295880141224598e-05,
+      "loss": 3.7748,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295041546473546e-05,
+      "loss": 3.7646,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294202951722494e-05,
+      "loss": 3.745,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293364356971442e-05,
+      "loss": 3.7528,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29252576222039e-05,
+      "loss": 3.7602,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916888053497106e-05,
+      "loss": 3.7644,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908502105986586e-05,
+      "loss": 3.763,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2900116158476066e-05,
+      "loss": 3.7577,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2891730210965546e-05,
+      "loss": 3.7447,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2883360642258755e-05,
+      "loss": 3.7539,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2874974694748235e-05,
+      "loss": 3.7504,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866588747237715e-05,
+      "loss": 3.7558,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2858202799727195e-05,
+      "loss": 3.7393,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284983323102041e-05,
+      "loss": 3.7707,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284146366231362e-05,
+      "loss": 3.7493,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28330777148031e-05,
+      "loss": 3.7579,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282469176729258e-05,
+      "loss": 3.7464,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281630581978206e-05,
+      "loss": 3.7564,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280791987227154e-05,
+      "loss": 3.754,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279953392476102e-05,
+      "loss": 3.7483,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27911479772505e-05,
+      "loss": 3.7585,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278276202973998e-05,
+      "loss": 3.7665,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277440883983692e-05,
+      "loss": 3.7506,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2766022892326405e-05,
+      "loss": 3.7613,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757636944815885e-05,
+      "loss": 3.7506,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749250997305365e-05,
+      "loss": 3.7621,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740865049794845e-05,
+      "loss": 3.7429,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2732479102284325e-05,
+      "loss": 3.7614,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2724093154773805e-05,
+      "loss": 3.7452,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715707207263285e-05,
+      "loss": 3.7586,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2707337638556494e-05,
+      "loss": 3.751,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698951691045974e-05,
+      "loss": 3.7521,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690565743535454e-05,
+      "loss": 3.7375,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268217979602493e-05,
+      "loss": 3.75,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267381022731814e-05,
+      "loss": 3.7506,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266542427980762e-05,
+      "loss": 3.7577,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26570383322971e-05,
+      "loss": 3.7494,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264865238478658e-05,
+      "loss": 3.7594,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26402828160798e-05,
+      "loss": 3.7438,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263189686856928e-05,
+      "loss": 3.7528,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262351092105876e-05,
+      "loss": 3.7586,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261512497354823e-05,
+      "loss": 3.7602,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260675540484145e-05,
+      "loss": 3.7506,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259836945733093e-05,
+      "loss": 3.7527,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25899835098204e-05,
+      "loss": 3.7455,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258159756230988e-05,
+      "loss": 3.7652,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2573227993603096e-05,
+      "loss": 3.743,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564842046092576e-05,
+      "loss": 3.756,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2556456098582056e-05,
+      "loss": 3.745,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2548070151071536e-05,
+      "loss": 3.7493,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253970058236475e-05,
+      "loss": 3.7614,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253131463485423e-05,
+      "loss": 3.7493,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522928687343705e-05,
+      "loss": 3.7501,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514542739833185e-05,
+      "loss": 3.7355,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25061731711264e-05,
+      "loss": 3.7495,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.872626304626465,
+      "eval_runtime": 301.7778,
+      "eval_samples_per_second": 1264.477,
+      "eval_steps_per_second": 39.516,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497787223615874e-05,
+      "loss": 3.7408,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2489401276105354e-05,
+      "loss": 3.7379,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2481015328594834e-05,
+      "loss": 3.7558,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472629381084314e-05,
+      "loss": 3.7527,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464243433573794e-05,
+      "loss": 3.7626,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245585748606328e-05,
+      "loss": 3.7433,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244747153855276e-05,
+      "loss": 3.7428,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243908559104224e-05,
+      "loss": 3.7417,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243071602233545e-05,
+      "loss": 3.7402,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242233007482493e-05,
+      "loss": 3.7516,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241394412731441e-05,
+      "loss": 3.7438,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240555817980389e-05,
+      "loss": 3.7498,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239720498990083e-05,
+      "loss": 3.7423,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238881904239031e-05,
+      "loss": 3.7359,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238043309487979e-05,
+      "loss": 3.7361,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237204714736927e-05,
+      "loss": 3.7331,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236366119985875e-05,
+      "loss": 3.7468,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2355275252348234e-05,
+      "loss": 3.7425,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346889304837714e-05,
+      "loss": 3.7416,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2338503357327194e-05,
+      "loss": 3.7582,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23301337886204e-05,
+      "loss": 3.7524,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232174784110988e-05,
+      "loss": 3.739,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231336189359936e-05,
+      "loss": 3.7376,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230497594608884e-05,
+      "loss": 3.7448,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229660637738205e-05,
+      "loss": 3.739,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228822042987153e-05,
+      "loss": 3.7423,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227983448236101e-05,
+      "loss": 3.7412,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227144853485049e-05,
+      "loss": 3.7397,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22630789661437e-05,
+      "loss": 3.7335,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225469301863319e-05,
+      "loss": 3.7298,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224630707112267e-05,
+      "loss": 3.7296,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223792112361215e-05,
+      "loss": 3.7393,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222955155490536e-05,
+      "loss": 3.7479,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222116560739484e-05,
+      "loss": 3.7456,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221277965988432e-05,
+      "loss": 3.728,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2204410091177526e-05,
+      "loss": 3.7397,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2196024143667006e-05,
+      "loss": 3.7423,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187638196156486e-05,
+      "loss": 3.7224,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2179252248645966e-05,
+      "loss": 3.7325,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170882679939175e-05,
+      "loss": 3.7315,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2162496732428655e-05,
+      "loss": 3.7239,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2154110784918135e-05,
+      "loss": 3.7442,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214572483740762e-05,
+      "loss": 3.7334,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213735526870083e-05,
+      "loss": 3.737,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212896932119031e-05,
+      "loss": 3.7338,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212058337367979e-05,
+      "loss": 3.7335,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211219742616927e-05,
+      "loss": 3.7413,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210382785746248e-05,
+      "loss": 3.7231,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209544190995196e-05,
+      "loss": 3.7014,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208705596244144e-05,
+      "loss": 3.7487,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207867001493092e-05,
+      "loss": 3.7287,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207030044622413e-05,
+      "loss": 3.7378,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206191449871361e-05,
+      "loss": 3.7245,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205352855120309e-05,
+      "loss": 3.7249,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2045142603692575e-05,
+      "loss": 3.7176,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036773034985784e-05,
+      "loss": 3.73,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028387087475264e-05,
+      "loss": 3.7345,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2020001139964744e-05,
+      "loss": 3.7203,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011615192454224e-05,
+      "loss": 3.7411,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200324562374743e-05,
+      "loss": 3.7242,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199485967623691e-05,
+      "loss": 3.7128,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198647372872639e-05,
+      "loss": 3.7302,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197808778121587e-05,
+      "loss": 3.7152,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196971821250908e-05,
+      "loss": 3.7092,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196133226499856e-05,
+      "loss": 3.7365,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195294631748804e-05,
+      "loss": 3.7333,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194456036997753e-05,
+      "loss": 3.7203,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193619080127074e-05,
+      "loss": 3.7253,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192780485376022e-05,
+      "loss": 3.7122,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191943528505343e-05,
+      "loss": 3.7282,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191104933754291e-05,
+      "loss": 3.7261,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190266339003239e-05,
+      "loss": 3.7303,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189427744252187e-05,
+      "loss": 3.7261,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188589149501135e-05,
+      "loss": 3.7304,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187750554750083e-05,
+      "loss": 3.7426,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18691195999903e-05,
+      "loss": 3.7146,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186073365247978e-05,
+      "loss": 3.7316,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185238046257673e-05,
+      "loss": 3.7255,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184399451506621e-05,
+      "loss": 3.7153,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183560856755569e-05,
+      "loss": 3.7336,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182722262004517e-05,
+      "loss": 3.7394,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181883667253465e-05,
+      "loss": 3.7282,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181045072502413e-05,
+      "loss": 3.7422,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1802064777513605e-05,
+      "loss": 3.7163,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793678830003085e-05,
+      "loss": 3.715,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17853092612963e-05,
+      "loss": 3.7244,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776923313785774e-05,
+      "loss": 3.7328,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1768537366275254e-05,
+      "loss": 3.7201,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760151418764734e-05,
+      "loss": 3.7242,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175178185005795e-05,
+      "loss": 3.7211,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174339590254743e-05,
+      "loss": 3.7299,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173500995503691e-05,
+      "loss": 3.717,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172662400752639e-05,
+      "loss": 3.718,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1718254438819605e-05,
+      "loss": 3.7247,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170986849130908e-05,
+      "loss": 3.7375,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170148254379856e-05,
+      "loss": 3.7263,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169309659628804e-05,
+      "loss": 3.7066,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168472702758125e-05,
+      "loss": 3.7202,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167634108007073e-05,
+      "loss": 3.7198,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166795513256021e-05,
+      "loss": 3.729,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165956918504969e-05,
+      "loss": 3.7235,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16511996163429e-05,
+      "loss": 3.7203,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164281366883238e-05,
+      "loss": 3.7101,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163442772132186e-05,
+      "loss": 3.713,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162604177381134e-05,
+      "loss": 3.7214,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161767220510455e-05,
+      "loss": 3.7156,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160928625759403e-05,
+      "loss": 3.7038,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160090031008351e-05,
+      "loss": 3.7344,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159253074137672e-05,
+      "loss": 3.7123,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15841447938662e-05,
+      "loss": 3.7178,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157575884635568e-05,
+      "loss": 3.7123,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156737289884516e-05,
+      "loss": 3.7179,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155898695133464e-05,
+      "loss": 3.7199,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155060100382413e-05,
+      "loss": 3.7124,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154221505631361e-05,
+      "loss": 3.722,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153382910880309e-05,
+      "loss": 3.7244,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1525475918900026e-05,
+      "loss": 3.7199,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1517089971389506e-05,
+      "loss": 3.7248,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508704023878986e-05,
+      "loss": 3.7132,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1500318076368466e-05,
+      "loss": 3.7239,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1491932128857946e-05,
+      "loss": 3.7095,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1483546181347426e-05,
+      "loss": 3.7235,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1475160233836906e-05,
+      "loss": 3.7098,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466774286326386e-05,
+      "loss": 3.7221,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1458404717619595e-05,
+      "loss": 3.7188,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145001877010908e-05,
+      "loss": 3.7158,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144163282259856e-05,
+      "loss": 3.7049,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143324687508804e-05,
+      "loss": 3.7113,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142487730638125e-05,
+      "loss": 3.7106,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141649135887073e-05,
+      "loss": 3.7234,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140810541136021e-05,
+      "loss": 3.7105,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139971946384969e-05,
+      "loss": 3.7253,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13913498951429e-05,
+      "loss": 3.7052,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138296394763238e-05,
+      "loss": 3.7186,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137457800012186e-05,
+      "loss": 3.7221,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136619205261134e-05,
+      "loss": 3.7247,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135782248390455e-05,
+      "loss": 3.7177,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349436536394035e-05,
+      "loss": 3.7167,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1341050588883515e-05,
+      "loss": 3.709,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332664641372995e-05,
+      "loss": 3.7287,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1324295072666204e-05,
+      "loss": 3.706,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315909125155684e-05,
+      "loss": 3.7195,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307523177645164e-05,
+      "loss": 3.7099,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1299137230134644e-05,
+      "loss": 3.7198,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129076766142785e-05,
+      "loss": 3.7172,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128238171391733e-05,
+      "loss": 3.7131,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127399576640681e-05,
+      "loss": 3.7151,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126560981889629e-05,
+      "loss": 3.7022,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125722387138577e-05,
+      "loss": 3.7173,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8620269298553467,
+      "eval_runtime": 321.06,
+      "eval_samples_per_second": 1188.535,
+      "eval_steps_per_second": 37.143,
+      "step": 534240
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.70281162547925e+17,
+  "trial_name": null,
+  "trial_params": null
+}