diff --git "a/checkpoint-2900160/trainer_state.json" "b/checkpoint-2900160/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2900160/trainer_state.json"
@@ -0,0 +1,34313 @@
+{
+  "best_metric": 3.8400497436523438,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full2/transformer/2/checkpoints/checkpoint-1068480",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 2900160,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9422,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8087,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1703,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9608,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.7996,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6916,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5901,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.523,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4358,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3874,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3369,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.314,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2569,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.1946,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1689,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1321,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1015,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0742,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0484,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0067,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0017,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9672,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9451,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9183,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9163,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8894,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8614,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8452,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8264,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8155,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7933,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7857,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7723,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7592,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7553,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7416,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7268,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7002,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.6953,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6778,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6755,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.66,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6553,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6391,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6446,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.628,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6295,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6205,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5924,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5849,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.5966,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5827,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5687,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5534,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5586,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.541,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5533,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5209,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5315,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5306,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5065,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5231,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.4931,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4943,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4831,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.499,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4698,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4754,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4598,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4567,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4699,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4622,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4519,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4515,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4598,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4314,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4479,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4221,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4205,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4192,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4218,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4141,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4203,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930422841748654e-05,
+      "loss": 4.405,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3855,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.3888,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.3976,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.3846,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92623314375414e-05,
+      "loss": 4.384,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3887,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3706,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3766,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9228804026303046e-05,
+      "loss": 4.3697,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922041807879253e-05,
+      "loss": 4.3729,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921203213128201e-05,
+      "loss": 4.3717,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3673,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3456,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3532,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3469,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3496,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3527,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3321,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3376,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3388,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.331,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3203,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3241,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3132,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3117,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3103,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3093,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3054,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3098,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3007,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3037,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3106,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.3055,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.3007,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2906,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.2898,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2885,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.276,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2818,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2757,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2805,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2717,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2796,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2732,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2653,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2604,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2761,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2671,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2593,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.261,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2641,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2588,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2614,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2507,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2497,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.2491,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2409,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.2378,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2372,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2402,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.231,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2402,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2315,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2336,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2379,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.22,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.190673351287842,
+      "eval_runtime": 313.8813,
+      "eval_samples_per_second": 1215.718,
+      "eval_steps_per_second": 37.992,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2087,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2056,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2306,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.213,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2179,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2089,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.2014,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.1951,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2024,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.195,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.2182,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.204,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.1856,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.1869,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.1886,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.1748,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.1901,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.1849,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1734,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.1979,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1772,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1787,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.1701,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1879,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1655,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.1663,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.1633,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.1611,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1591,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1589,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.1608,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.1613,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1582,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1634,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845757529499864e-05,
+      "loss": 4.1562,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844918934748812e-05,
+      "loss": 4.156,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1427,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1452,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1361,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841566193624977e-05,
+      "loss": 4.1428,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1361,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1453,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1273,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.145,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.1376,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.1396,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1357,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.119,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1209,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.1367,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1266,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1237,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83066937550242e-05,
+      "loss": 4.1132,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829832418631741e-05,
+      "loss": 4.1247,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828993823880689e-05,
+      "loss": 4.1041,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828155229129637e-05,
+      "loss": 4.1307,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.1008,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1183,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.1165,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1029,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.1181,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231252985036985e-05,
+      "loss": 4.1046,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222867037526465e-05,
+      "loss": 4.1037,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.0921,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820609514250542e-05,
+      "loss": 4.1169,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197725573798634e-05,
+      "loss": 4.09,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189339626288114e-05,
+      "loss": 4.1017,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.0822,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.095,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816419816256029e-05,
+      "loss": 4.1034,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815581221504977e-05,
+      "loss": 4.1029,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.0965,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813904032002872e-05,
+      "loss": 4.1001,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813067075132194e-05,
+      "loss": 4.1053,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812228480381141e-05,
+      "loss": 4.0903,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.0999,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810551290879037e-05,
+      "loss": 4.0867,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809714334008359e-05,
+      "loss": 4.0878,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808875739257306e-05,
+      "loss": 4.0841,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808037144506255e-05,
+      "loss": 4.0923,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807198549755203e-05,
+      "loss": 4.0923,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063615928845243e-05,
+      "loss": 4.0955,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055229981334717e-05,
+      "loss": 4.0859,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046844033824197e-05,
+      "loss": 4.0665,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8038458086313677e-05,
+      "loss": 4.0713,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030088517606886e-05,
+      "loss": 4.0825,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021702570096366e-05,
+      "loss": 4.0759,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801333300138958e-05,
+      "loss": 4.0736,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800494705387906e-05,
+      "loss": 4.0844,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996561106368535e-05,
+      "loss": 4.0675,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988175158858015e-05,
+      "loss": 4.0709,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979789211347494e-05,
+      "loss": 4.0735,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797140326383698e-05,
+      "loss": 4.0799,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796301731632646e-05,
+      "loss": 4.0779,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795464774761967e-05,
+      "loss": 4.0819,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794626180010915e-05,
+      "loss": 4.057,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793787585259863e-05,
+      "loss": 4.0662,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792948990508811e-05,
+      "loss": 4.0647,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792110395757759e-05,
+      "loss": 4.0667,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791271801006707e-05,
+      "loss": 4.0775,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.0567,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789594611504603e-05,
+      "loss": 4.0598,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788757654633924e-05,
+      "loss": 4.0676,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787919059882872e-05,
+      "loss": 4.0645,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708046513182e-05,
+      "loss": 4.0471,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862435082611415e-05,
+      "loss": 4.0611,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854049135100895e-05,
+      "loss": 4.0514,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845663187590375e-05,
+      "loss": 4.0493,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837277240079855e-05,
+      "loss": 4.0486,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828891292569335e-05,
+      "loss": 4.0494,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820505345058815e-05,
+      "loss": 4.0466,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812119397548295e-05,
+      "loss": 4.0576,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803733450037775e-05,
+      "loss": 4.0434,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795363881330984e-05,
+      "loss": 4.0529,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786977933820464e-05,
+      "loss": 4.0626,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778591986309944e-05,
+      "loss": 4.0558,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770206038799423e-05,
+      "loss": 4.0535,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0486,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0447,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450645750716e-05,
+      "loss": 4.0469,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773667862756108e-05,
+      "loss": 4.0384,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772830905885429e-05,
+      "loss": 4.0395,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771992311134377e-05,
+      "loss": 4.0396,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771153716383325e-05,
+      "loss": 4.045,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770315121632272e-05,
+      "loss": 4.0415,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76947652688122e-05,
+      "loss": 4.0417,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768637932130168e-05,
+      "loss": 4.0424,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767799337379116e-05,
+      "loss": 4.0324,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766962380508437e-05,
+      "loss": 4.0323,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766123785757386e-05,
+      "loss": 4.0522,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765285191006334e-05,
+      "loss": 4.0438,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764446596255282e-05,
+      "loss": 4.0392,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636096393846026e-05,
+      "loss": 4.0335,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7627710446335506e-05,
+      "loss": 4.0396,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7619324498824986e-05,
+      "loss": 4.0358,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610938551314466e-05,
+      "loss": 4.0433,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602568982607675e-05,
+      "loss": 4.0352,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594183035097155e-05,
+      "loss": 4.0313,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585797087586635e-05,
+      "loss": 4.032,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577411140076115e-05,
+      "loss": 4.03,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569025192565595e-05,
+      "loss": 4.0221,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560639245055075e-05,
+      "loss": 4.0282,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552253297544555e-05,
+      "loss": 4.0327,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754388372883777e-05,
+      "loss": 4.0172,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753549778132725e-05,
+      "loss": 4.0351,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752711183381673e-05,
+      "loss": 4.0276,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751872588630621e-05,
+      "loss": 4.0239,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751035631759942e-05,
+      "loss": 4.0298,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75019703700889e-05,
+      "loss": 4.0167,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.023011207580566,
+      "eval_runtime": 312.5224,
+      "eval_samples_per_second": 1221.004,
+      "eval_steps_per_second": 38.157,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749358442257838e-05,
+      "loss": 4.0052,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748519847506786e-05,
+      "loss": 4.0054,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747681252755734e-05,
+      "loss": 4.03,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746842658004682e-05,
+      "loss": 4.0141,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74600406325363e-05,
+      "loss": 4.0255,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745165468502578e-05,
+      "loss": 4.0098,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744326873751526e-05,
+      "loss": 4.0121,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743488279000474e-05,
+      "loss": 4.001,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426513221297955e-05,
+      "loss": 4.0119,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418127273787435e-05,
+      "loss": 4.0051,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740974132627691e-05,
+      "loss": 4.025,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740135537876639e-05,
+      "loss": 4.0164,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739296943125587e-05,
+      "loss": 3.9958,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738458348374535e-05,
+      "loss": 3.9997,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737619753623483e-05,
+      "loss": 4.0062,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736781158872431e-05,
+      "loss": 3.988,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735944202001752e-05,
+      "loss": 4.0018,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7351056072507e-05,
+      "loss": 4.0036,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734267012499648e-05,
+      "loss": 3.9936,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733428417748596e-05,
+      "loss": 4.0194,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732591460877917e-05,
+      "loss": 3.9968,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731752866126865e-05,
+      "loss": 4.0034,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730915909256186e-05,
+      "loss": 3.9946,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730077314505134e-05,
+      "loss": 4.0106,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729238719754082e-05,
+      "loss": 3.9856,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72840012500303e-05,
+      "loss": 3.9942,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727561530251978e-05,
+      "loss": 3.9862,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726724573381299e-05,
+      "loss": 3.9915,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725885978630247e-05,
+      "loss": 3.9862,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725047383879195e-05,
+      "loss": 3.9859,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724208789128143e-05,
+      "loss": 3.9928,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723370194377091e-05,
+      "loss": 3.9938,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72253159962604e-05,
+      "loss": 3.9863,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721693004874988e-05,
+      "loss": 4.0005,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720854410123936e-05,
+      "loss": 3.986,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200174532532566e-05,
+      "loss": 3.9925,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191788585022046e-05,
+      "loss": 3.9784,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183402637511526e-05,
+      "loss": 3.9815,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175016690001006e-05,
+      "loss": 3.9703,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166630742490486e-05,
+      "loss": 3.9847,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158244794979966e-05,
+      "loss": 3.9737,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149858847469446e-05,
+      "loss": 3.9846,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141489278762655e-05,
+      "loss": 3.9709,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133103331252135e-05,
+      "loss": 3.9813,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124717383741615e-05,
+      "loss": 3.9846,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116331436231095e-05,
+      "loss": 3.9802,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710796186752431e-05,
+      "loss": 3.9817,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709957592001379e-05,
+      "loss": 3.9647,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091206351307e-05,
+      "loss": 3.9666,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708282040379648e-05,
+      "loss": 3.9826,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707443445628596e-05,
+      "loss": 3.9722,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706604850877544e-05,
+      "loss": 3.974,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705766256126492e-05,
+      "loss": 3.9598,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704929299255813e-05,
+      "loss": 3.9746,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704090704504761e-05,
+      "loss": 3.9505,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703252109753709e-05,
+      "loss": 3.9831,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702413515002657e-05,
+      "loss": 3.9506,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701574920251605e-05,
+      "loss": 3.969,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700736325500553e-05,
+      "loss": 3.9726,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699897730749501e-05,
+      "loss": 3.9551,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699059135998449e-05,
+      "loss": 3.975,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69822217912777e-05,
+      "loss": 3.9565,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697383584376718e-05,
+      "loss": 3.9572,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696544989625666e-05,
+      "loss": 3.9472,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695708032754987e-05,
+      "loss": 3.971,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694869438003935e-05,
+      "loss": 3.9508,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694030843252883e-05,
+      "loss": 3.9576,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931922485018307e-05,
+      "loss": 3.9429,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923536537507787e-05,
+      "loss": 3.9508,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915150589997267e-05,
+      "loss": 3.9605,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690676464248675e-05,
+      "loss": 3.9654,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689837869497623e-05,
+      "loss": 3.9527,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689000912626944e-05,
+      "loss": 3.9661,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688162317875892e-05,
+      "loss": 3.9664,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68732372312484e-05,
+      "loss": 3.9546,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686486766254161e-05,
+      "loss": 3.9569,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685648171503109e-05,
+      "loss": 3.9555,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684809576752057e-05,
+      "loss": 3.9497,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683970982001005e-05,
+      "loss": 3.949,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683132387249953e-05,
+      "loss": 3.9566,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682295430379274e-05,
+      "loss": 3.9595,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681456835628222e-05,
+      "loss": 3.9631,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680618240877171e-05,
+      "loss": 3.9496,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679779646126119e-05,
+      "loss": 3.9301,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678941051375067e-05,
+      "loss": 3.9395,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6781040945043876e-05,
+      "loss": 3.9489,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772654997533356e-05,
+      "loss": 3.9453,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764269050022836e-05,
+      "loss": 3.9424,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755883102512316e-05,
+      "loss": 3.9509,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747513533805525e-05,
+      "loss": 3.9427,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739127586295005e-05,
+      "loss": 3.9362,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730741638784485e-05,
+      "loss": 3.9441,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722355691273965e-05,
+      "loss": 3.9518,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713969743763445e-05,
+      "loss": 3.9476,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670560017505666e-05,
+      "loss": 3.9535,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669721422754614e-05,
+      "loss": 3.9364,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668882828003562e-05,
+      "loss": 3.9395,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66804423325251e-05,
+      "loss": 3.938,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667205638501458e-05,
+      "loss": 3.942,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663670437504054e-05,
+      "loss": 3.9494,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665530086879727e-05,
+      "loss": 3.9347,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664691492128675e-05,
+      "loss": 3.9348,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663852897377623e-05,
+      "loss": 3.9409,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66301430262657e-05,
+      "loss": 3.9433,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662175707875518e-05,
+      "loss": 3.9236,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66133875100484e-05,
+      "loss": 3.9345,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660500156253788e-05,
+      "loss": 3.9347,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659661561502736e-05,
+      "loss": 3.9269,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658822966751684e-05,
+      "loss": 3.9258,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657984372000632e-05,
+      "loss": 3.9289,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65714577724958e-05,
+      "loss": 3.9307,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656307182498528e-05,
+      "loss": 3.9347,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655470225627849e-05,
+      "loss": 3.9235,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654631630876797e-05,
+      "loss": 3.9337,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653793036125745e-05,
+      "loss": 3.9421,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652954441374693e-05,
+      "loss": 3.9428,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652115846623641e-05,
+      "loss": 3.9336,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512788897529616e-05,
+      "loss": 3.932,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504402950019096e-05,
+      "loss": 3.9278,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649601700250858e-05,
+      "loss": 3.9309,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648763105499806e-05,
+      "loss": 3.9189,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647926148629127e-05,
+      "loss": 3.9227,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647087553878075e-05,
+      "loss": 3.9234,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646248959127023e-05,
+      "loss": 3.9323,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645410364375971e-05,
+      "loss": 3.9263,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644571769624919e-05,
+      "loss": 3.9282,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643733174873867e-05,
+      "loss": 3.9274,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642894580122815e-05,
+      "loss": 3.9185,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642057623252136e-05,
+      "loss": 3.9173,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641219028501084e-05,
+      "loss": 3.9378,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640380433750032e-05,
+      "loss": 3.9348,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63954183899898e-05,
+      "loss": 3.925,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638703244247928e-05,
+      "loss": 3.9213,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378662873772496e-05,
+      "loss": 3.9245,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370276926261976e-05,
+      "loss": 3.9262,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361890978751456e-05,
+      "loss": 3.9369,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353505031240936e-05,
+      "loss": 3.923,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345135462534145e-05,
+      "loss": 3.9197,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336749515023625e-05,
+      "loss": 3.9197,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328363567513105e-05,
+      "loss": 3.9183,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319977620002585e-05,
+      "loss": 3.9162,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631159167249206e-05,
+      "loss": 3.9188,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303222103785274e-05,
+      "loss": 3.9222,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294836156274754e-05,
+      "loss": 3.9125,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286450208764234e-05,
+      "loss": 3.924,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278064261253714e-05,
+      "loss": 3.9238,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626969469254693e-05,
+      "loss": 3.9154,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626130874503641e-05,
+      "loss": 3.9216,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625292279752589e-05,
+      "loss": 3.9066,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.948953628540039,
+      "eval_runtime": 326.0308,
+      "eval_samples_per_second": 1170.414,
+      "eval_steps_per_second": 36.576,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624453685001536e-05,
+      "loss": 3.9091,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623616728130858e-05,
+      "loss": 3.8996,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622778133379806e-05,
+      "loss": 3.924,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621939538628753e-05,
+      "loss": 3.9085,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621100943877701e-05,
+      "loss": 3.925,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620262349126649e-05,
+      "loss": 3.9025,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619423754375597e-05,
+      "loss": 3.9116,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618586797504919e-05,
+      "loss": 3.8936,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617748202753867e-05,
+      "loss": 3.9116,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616909608002815e-05,
+      "loss": 3.8989,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616071013251763e-05,
+      "loss": 3.9217,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615232418500711e-05,
+      "loss": 3.9142,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614395461630032e-05,
+      "loss": 3.8917,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61355686687898e-05,
+      "loss": 3.8997,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612718272127928e-05,
+      "loss": 3.9063,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611879677376876e-05,
+      "loss": 3.8902,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110427205061966e-05,
+      "loss": 3.9008,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6102041257551446e-05,
+      "loss": 3.9027,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093655310040926e-05,
+      "loss": 3.8941,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085269362530406e-05,
+      "loss": 3.9224,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607689979382362e-05,
+      "loss": 3.8993,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60685138463131e-05,
+      "loss": 3.9016,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606012789880258e-05,
+      "loss": 3.9008,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605174195129206e-05,
+      "loss": 3.9064,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604335600378154e-05,
+      "loss": 3.8899,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603498643507475e-05,
+      "loss": 3.8974,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602660048756423e-05,
+      "loss": 3.8922,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601821454005371e-05,
+      "loss": 3.8945,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600982859254319e-05,
+      "loss": 3.888,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600144264503267e-05,
+      "loss": 3.8909,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599307307632588e-05,
+      "loss": 3.8924,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598468712881536e-05,
+      "loss": 3.899,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976301181304846e-05,
+      "loss": 3.899,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967915233794326e-05,
+      "loss": 3.9,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959529286283806e-05,
+      "loss": 3.8922,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951159717577015e-05,
+      "loss": 3.8962,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942773770066495e-05,
+      "loss": 3.8798,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934387822555975e-05,
+      "loss": 3.894,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5926001875045455e-05,
+      "loss": 3.8771,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917615927534935e-05,
+      "loss": 3.8864,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909246358828144e-05,
+      "loss": 3.8805,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900860411317624e-05,
+      "loss": 3.8958,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892474463807104e-05,
+      "loss": 3.8786,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884088516296584e-05,
+      "loss": 3.8847,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875702568786064e-05,
+      "loss": 3.8955,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867316621275544e-05,
+      "loss": 3.8892,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585894705256876e-05,
+      "loss": 3.8921,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585056110505824e-05,
+      "loss": 3.8746,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584217515754772e-05,
+      "loss": 3.8748,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58337892100372e-05,
+      "loss": 3.8901,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582541964133041e-05,
+      "loss": 3.8844,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581703369381989e-05,
+      "loss": 3.8804,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580864774630937e-05,
+      "loss": 3.8725,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580026179879884e-05,
+      "loss": 3.8848,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579189223009206e-05,
+      "loss": 3.8638,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578350628258154e-05,
+      "loss": 3.8851,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577512033507102e-05,
+      "loss": 3.8671,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57667343875605e-05,
+      "loss": 3.8762,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575836481885371e-05,
+      "loss": 3.8866,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574997887134319e-05,
+      "loss": 3.8671,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741592923832666e-05,
+      "loss": 3.8825,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733206976322146e-05,
+      "loss": 3.8712,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724821028811626e-05,
+      "loss": 3.8684,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571645146010484e-05,
+      "loss": 3.8653,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708065512594315e-05,
+      "loss": 3.8806,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699679565083795e-05,
+      "loss": 3.8644,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691293617573275e-05,
+      "loss": 3.8713,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568292404886649e-05,
+      "loss": 3.8579,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567453810135597e-05,
+      "loss": 3.863,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566615215384545e-05,
+      "loss": 3.8748,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565776620633493e-05,
+      "loss": 3.8821,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564939663762814e-05,
+      "loss": 3.8674,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564101069011762e-05,
+      "loss": 3.8812,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56326247426071e-05,
+      "loss": 3.8815,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562423879509658e-05,
+      "loss": 3.8703,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561586922638979e-05,
+      "loss": 3.875,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560748327887927e-05,
+      "loss": 3.8699,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559909733136875e-05,
+      "loss": 3.8672,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559071138385823e-05,
+      "loss": 3.8664,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582341815151445e-05,
+      "loss": 3.8733,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5573955867640925e-05,
+      "loss": 3.8775,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565569920130405e-05,
+      "loss": 3.871,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557183972619885e-05,
+      "loss": 3.8703,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548814403913094e-05,
+      "loss": 3.8498,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540428456402574e-05,
+      "loss": 3.8551,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5532042508892054e-05,
+      "loss": 3.8646,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523656561381534e-05,
+      "loss": 3.8648,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515270613871014e-05,
+      "loss": 3.8609,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550690104516422e-05,
+      "loss": 3.8683,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54985150976537e-05,
+      "loss": 3.862,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549012915014318e-05,
+      "loss": 3.8556,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548174320263266e-05,
+      "loss": 3.8612,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547337363392588e-05,
+      "loss": 3.8677,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546498768641536e-05,
+      "loss": 3.8661,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545660173890484e-05,
+      "loss": 3.8755,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544821579139432e-05,
+      "loss": 3.857,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543984622268753e-05,
+      "loss": 3.8562,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543146027517701e-05,
+      "loss": 3.8589,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542307432766649e-05,
+      "loss": 3.8633,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541468838015597e-05,
+      "loss": 3.8695,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406318811449176e-05,
+      "loss": 3.8553,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397932863938656e-05,
+      "loss": 3.8563,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389546916428136e-05,
+      "loss": 3.8606,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381160968917616e-05,
+      "loss": 3.8647,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537279140021083e-05,
+      "loss": 3.8472,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536440545270031e-05,
+      "loss": 3.8569,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535601950518979e-05,
+      "loss": 3.8565,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534763355767927e-05,
+      "loss": 3.8481,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533926398897248e-05,
+      "loss": 3.8459,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533087804146196e-05,
+      "loss": 3.8532,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532249209395144e-05,
+      "loss": 3.8555,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531410614644092e-05,
+      "loss": 3.8542,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530573657773413e-05,
+      "loss": 3.8495,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529735063022361e-05,
+      "loss": 3.8512,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528896468271309e-05,
+      "loss": 3.8677,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528057873520257e-05,
+      "loss": 3.8643,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272209166495786e-05,
+      "loss": 3.859,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263823218985266e-05,
+      "loss": 3.8558,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255437271474746e-05,
+      "loss": 3.8515,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247051323964226e-05,
+      "loss": 3.8515,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238665376453706e-05,
+      "loss": 3.8438,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230312186550644e-05,
+      "loss": 3.8475,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221926239040124e-05,
+      "loss": 3.8488,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213540291529604e-05,
+      "loss": 3.8573,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5205154344019084e-05,
+      "loss": 3.8482,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196768396508564e-05,
+      "loss": 3.856,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188382448998044e-05,
+      "loss": 3.8493,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5179996501487524e-05,
+      "loss": 3.8462,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171610553977003e-05,
+      "loss": 3.8422,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516324098527022e-05,
+      "loss": 3.8617,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51548550377597e-05,
+      "loss": 3.8636,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514646909024918e-05,
+      "loss": 3.851,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513808314273866e-05,
+      "loss": 3.8447,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512971357403187e-05,
+      "loss": 3.8537,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512132762652135e-05,
+      "loss": 3.8485,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511294167901083e-05,
+      "loss": 3.8632,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510455573150031e-05,
+      "loss": 3.8459,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509618616279352e-05,
+      "loss": 3.8498,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087800215283e-05,
+      "loss": 3.8461,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507941426777248e-05,
+      "loss": 3.8518,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507102832026196e-05,
+      "loss": 3.8401,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506265875155517e-05,
+      "loss": 3.845,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505427280404465e-05,
+      "loss": 3.85,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504588685653413e-05,
+      "loss": 3.8432,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503750090902361e-05,
+      "loss": 3.8503,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502913134031682e-05,
+      "loss": 3.8537,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50207453928063e-05,
+      "loss": 3.8437,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501235944529578e-05,
+      "loss": 3.8468,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500397349778526e-05,
+      "loss": 3.8414,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9086923599243164,
+      "eval_runtime": 344.9677,
+      "eval_samples_per_second": 1106.164,
+      "eval_steps_per_second": 34.568,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499558755027474e-05,
+      "loss": 3.8296,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987201602764215e-05,
+      "loss": 3.8294,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978815655253695e-05,
+      "loss": 3.85,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970429707743175e-05,
+      "loss": 3.8429,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496204376023266e-05,
+      "loss": 3.8556,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495365781272214e-05,
+      "loss": 3.8316,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494527186521162e-05,
+      "loss": 3.8435,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49368859177011e-05,
+      "loss": 3.8248,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492851634899431e-05,
+      "loss": 3.8453,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492013040148379e-05,
+      "loss": 3.8296,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491174445397327e-05,
+      "loss": 3.847,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490337488526648e-05,
+      "loss": 3.847,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489498893775596e-05,
+      "loss": 3.8209,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488660299024544e-05,
+      "loss": 3.8352,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487821704273492e-05,
+      "loss": 3.8346,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48698310952244e-05,
+      "loss": 3.8221,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486146152651761e-05,
+      "loss": 3.8325,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853075579007095e-05,
+      "loss": 3.8339,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844689631496575e-05,
+      "loss": 3.8309,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836320062789784e-05,
+      "loss": 3.8499,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827934115279264e-05,
+      "loss": 3.8284,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819548167768744e-05,
+      "loss": 3.834,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4811162220258224e-05,
+      "loss": 3.8355,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480279265155143e-05,
+      "loss": 3.8371,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479440670404091e-05,
+      "loss": 3.8227,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478602075653039e-05,
+      "loss": 3.8356,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477763480901987e-05,
+      "loss": 3.8225,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476926524031308e-05,
+      "loss": 3.8255,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476087929280256e-05,
+      "loss": 3.8209,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475249334529205e-05,
+      "loss": 3.8246,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474410739778153e-05,
+      "loss": 3.8264,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473573782907474e-05,
+      "loss": 3.8337,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472735188156422e-05,
+      "loss": 3.8323,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47189659340537e-05,
+      "loss": 3.8355,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471057998654318e-05,
+      "loss": 3.8228,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470221041783639e-05,
+      "loss": 3.8328,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469382447032587e-05,
+      "loss": 3.8204,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468543852281535e-05,
+      "loss": 3.8212,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467705257530483e-05,
+      "loss": 3.814,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668683006598036e-05,
+      "loss": 3.8202,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660297059087516e-05,
+      "loss": 3.8192,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651911111577e-05,
+      "loss": 3.83,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464352516406648e-05,
+      "loss": 3.8146,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463515559535969e-05,
+      "loss": 3.821,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462676964784917e-05,
+      "loss": 3.835,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461838370033865e-05,
+      "loss": 3.8252,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460999775282813e-05,
+      "loss": 3.8274,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460162818412134e-05,
+      "loss": 3.8143,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459324223661082e-05,
+      "loss": 3.8045,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45848562891003e-05,
+      "loss": 3.8269,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457647034158978e-05,
+      "loss": 3.8225,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456810077288299e-05,
+      "loss": 3.814,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455971482537247e-05,
+      "loss": 3.8129,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551328877861956e-05,
+      "loss": 3.8218,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542942930351436e-05,
+      "loss": 3.8004,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534573361644645e-05,
+      "loss": 3.824,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526187414134125e-05,
+      "loss": 3.8035,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517801466623605e-05,
+      "loss": 3.8157,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509415519113085e-05,
+      "loss": 3.8199,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501045950406294e-05,
+      "loss": 3.8082,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492660002895774e-05,
+      "loss": 3.8156,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4484274055385254e-05,
+      "loss": 3.813,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475888107874734e-05,
+      "loss": 3.8039,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446751853916794e-05,
+      "loss": 3.8102,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445913259165742e-05,
+      "loss": 3.8143,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445074664414691e-05,
+      "loss": 3.8048,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444236069663639e-05,
+      "loss": 3.8095,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44339911279296e-05,
+      "loss": 3.7993,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442560518041908e-05,
+      "loss": 3.7994,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441721923290856e-05,
+      "loss": 3.8125,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440883328539804e-05,
+      "loss": 3.8226,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440046371669125e-05,
+      "loss": 3.8047,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439207776918073e-05,
+      "loss": 3.8233,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438369182167021e-05,
+      "loss": 3.8199,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437530587415969e-05,
+      "loss": 3.8079,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669363054529e-05,
+      "loss": 3.8102,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435855035794238e-05,
+      "loss": 3.8153,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4350164410431864e-05,
+      "loss": 3.8069,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341778462921344e-05,
+      "loss": 3.8056,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433340889421455e-05,
+      "loss": 3.8149,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432502294670403e-05,
+      "loss": 3.8126,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431663699919351e-05,
+      "loss": 3.8133,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430825105168299e-05,
+      "loss": 3.8073,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42998814829762e-05,
+      "loss": 3.7929,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429149553546568e-05,
+      "loss": 3.7939,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428310958795516e-05,
+      "loss": 3.8064,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427472364044464e-05,
+      "loss": 3.8088,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426635407173785e-05,
+      "loss": 3.8014,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425796812422733e-05,
+      "loss": 3.8056,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424958217671682e-05,
+      "loss": 3.8016,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42411962292063e-05,
+      "loss": 3.7988,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232826660499506e-05,
+      "loss": 3.8043,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224440712988986e-05,
+      "loss": 3.8052,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216054765478466e-05,
+      "loss": 3.8121,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207668817967946e-05,
+      "loss": 3.8119,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199299249261155e-05,
+      "loss": 3.8058,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190913301750635e-05,
+      "loss": 3.7948,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182527354240115e-05,
+      "loss": 3.8061,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174141406729595e-05,
+      "loss": 3.8029,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165771838022804e-05,
+      "loss": 3.8102,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157385890512284e-05,
+      "loss": 3.8005,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414899994300177e-05,
+      "loss": 3.7932,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140613995491244e-05,
+      "loss": 3.8082,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413224442678446e-05,
+      "loss": 3.8048,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412385847927394e-05,
+      "loss": 3.7959,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411547253176342e-05,
+      "loss": 3.8009,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410708658425289e-05,
+      "loss": 3.8005,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409871701554611e-05,
+      "loss": 3.7836,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409033106803559e-05,
+      "loss": 3.7897,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408194512052507e-05,
+      "loss": 3.8019,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407355917301454e-05,
+      "loss": 3.7956,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406518960430776e-05,
+      "loss": 3.7987,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405680365679724e-05,
+      "loss": 3.7929,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404841770928672e-05,
+      "loss": 3.794,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4040048140579934e-05,
+      "loss": 3.8134,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031662193069414e-05,
+      "loss": 3.8095,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4023276245558894e-05,
+      "loss": 3.803,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401489029804837e-05,
+      "loss": 3.8001,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400652072934158e-05,
+      "loss": 3.7932,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399815116063479e-05,
+      "loss": 3.7949,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398976521312427e-05,
+      "loss": 3.7893,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398137926561375e-05,
+      "loss": 3.7976,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397299331810323e-05,
+      "loss": 3.7928,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396460737059271e-05,
+      "loss": 3.8019,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395622142308219e-05,
+      "loss": 3.7937,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394783547557167e-05,
+      "loss": 3.7979,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393944952806115e-05,
+      "loss": 3.7957,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393107995935437e-05,
+      "loss": 3.7897,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392269401184384e-05,
+      "loss": 3.7916,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391430806433332e-05,
+      "loss": 3.8046,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39059221168228e-05,
+      "loss": 3.8098,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897552548116016e-05,
+      "loss": 3.8004,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388916660060549e-05,
+      "loss": 3.7895,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388078065309497e-05,
+      "loss": 3.8009,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387239470558445e-05,
+      "loss": 3.7929,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3864025136877665e-05,
+      "loss": 3.8066,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855639189367145e-05,
+      "loss": 3.7937,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847253241856625e-05,
+      "loss": 3.7943,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383888367314984e-05,
+      "loss": 3.7902,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830497725639314e-05,
+      "loss": 3.8001,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822111778128794e-05,
+      "loss": 3.7852,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813725830618274e-05,
+      "loss": 3.7918,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805339883107754e-05,
+      "loss": 3.7948,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796953935597234e-05,
+      "loss": 3.7889,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788567988086714e-05,
+      "loss": 3.7971,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378019841937992e-05,
+      "loss": 3.799,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37718124718694e-05,
+      "loss": 3.793,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376342652435888e-05,
+      "loss": 3.7892,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375504057684836e-05,
+      "loss": 3.7903,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8839023113250732,
+      "eval_runtime": 315.3546,
+      "eval_samples_per_second": 1210.038,
+      "eval_steps_per_second": 37.815,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374667100814158e-05,
+      "loss": 3.7874,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373828506063106e-05,
+      "loss": 3.773,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372989911312054e-05,
+      "loss": 3.7974,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372151316561002e-05,
+      "loss": 3.7905,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371314359690323e-05,
+      "loss": 3.7989,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370475764939271e-05,
+      "loss": 3.783,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369637170188219e-05,
+      "loss": 3.7854,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368798575437167e-05,
+      "loss": 3.7774,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367961618566488e-05,
+      "loss": 3.7868,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367123023815436e-05,
+      "loss": 3.7822,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366284429064384e-05,
+      "loss": 3.7917,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365445834313332e-05,
+      "loss": 3.7935,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364608877442653e-05,
+      "loss": 3.7692,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363770282691601e-05,
+      "loss": 3.7822,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362931687940549e-05,
+      "loss": 3.7846,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362093093189497e-05,
+      "loss": 3.7693,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361256136318818e-05,
+      "loss": 3.7791,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360417541567766e-05,
+      "loss": 3.7816,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359578946816714e-05,
+      "loss": 3.7796,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358740352065662e-05,
+      "loss": 3.7963,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357903395194983e-05,
+      "loss": 3.7806,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357064800443931e-05,
+      "loss": 3.7876,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356226205692879e-05,
+      "loss": 3.7819,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3553892488222006e-05,
+      "loss": 3.7848,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545506540711486e-05,
+      "loss": 3.7732,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3537120593200966e-05,
+      "loss": 3.7829,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528734645690446e-05,
+      "loss": 3.7772,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520348698179926e-05,
+      "loss": 3.7708,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511962750669406e-05,
+      "loss": 3.7705,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503576803158886e-05,
+      "loss": 3.7763,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3495190855648366e-05,
+      "loss": 3.7764,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486821286941575e-05,
+      "loss": 3.7826,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478435339431055e-05,
+      "loss": 3.7839,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3470049391920535e-05,
+      "loss": 3.7834,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461663444410015e-05,
+      "loss": 3.7746,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3453293875703224e-05,
+      "loss": 3.7804,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344490792819271e-05,
+      "loss": 3.7728,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343652198068219e-05,
+      "loss": 3.7727,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342813603317167e-05,
+      "loss": 3.7622,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341976646446488e-05,
+      "loss": 3.7671,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341138051695436e-05,
+      "loss": 3.7715,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340299456944384e-05,
+      "loss": 3.7789,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339460862193332e-05,
+      "loss": 3.7673,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338623905322653e-05,
+      "loss": 3.7718,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337785310571601e-05,
+      "loss": 3.7859,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336946715820549e-05,
+      "loss": 3.7754,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336108121069497e-05,
+      "loss": 3.7789,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335269526318445e-05,
+      "loss": 3.7666,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334430931567393e-05,
+      "loss": 3.7546,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333592336816341e-05,
+      "loss": 3.7772,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327553799456624e-05,
+      "loss": 3.7756,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33191678519461e-05,
+      "loss": 3.7657,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331078190443558e-05,
+      "loss": 3.7646,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330239595692506e-05,
+      "loss": 3.7737,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329401000941454e-05,
+      "loss": 3.7467,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328562406190402e-05,
+      "loss": 3.7768,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32772381143935e-05,
+      "loss": 3.7584,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326885216688298e-05,
+      "loss": 3.764,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260482598176186e-05,
+      "loss": 3.7723,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3252096650665666e-05,
+      "loss": 3.764,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243710703155146e-05,
+      "loss": 3.7639,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323534113444836e-05,
+      "loss": 3.7683,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322695518693784e-05,
+      "loss": 3.7566,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321856923942732e-05,
+      "loss": 3.7569,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32101832919168e-05,
+      "loss": 3.7667,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320179734440628e-05,
+      "loss": 3.763,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319341139689576e-05,
+      "loss": 3.758,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318502544938524e-05,
+      "loss": 3.7547,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317663950187472e-05,
+      "loss": 3.7521,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316826993316793e-05,
+      "loss": 3.76,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315988398565741e-05,
+      "loss": 3.7722,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315151441695062e-05,
+      "loss": 3.7591,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31431284694401e-05,
+      "loss": 3.7757,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313474252192958e-05,
+      "loss": 3.7749,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3126356574419067e-05,
+      "loss": 3.7647,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3117970626908546e-05,
+      "loss": 3.7591,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109601058201756e-05,
+      "loss": 3.77,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3101215110691236e-05,
+      "loss": 3.7618,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3092829163180715e-05,
+      "loss": 3.7556,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3084443215670195e-05,
+      "loss": 3.7694,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3076057268159675e-05,
+      "loss": 3.764,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067671320649155e-05,
+      "loss": 3.7691,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305928537313863e-05,
+      "loss": 3.7623,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305089942562811e-05,
+      "loss": 3.7496,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042529856921324e-05,
+      "loss": 3.7446,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034143909410804e-05,
+      "loss": 3.7581,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302577434070402e-05,
+      "loss": 3.7638,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30173883931935e-05,
+      "loss": 3.7563,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300900244568298e-05,
+      "loss": 3.7581,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300061649817245e-05,
+      "loss": 3.7587,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299223055066193e-05,
+      "loss": 3.748,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298384460315141e-05,
+      "loss": 3.7627,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297545865564089e-05,
+      "loss": 3.7534,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296707270813037e-05,
+      "loss": 3.7669,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295871951822732e-05,
+      "loss": 3.7699,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29503335707168e-05,
+      "loss": 3.7618,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294194762320628e-05,
+      "loss": 3.7517,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293356167569576e-05,
+      "loss": 3.7545,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292517572818524e-05,
+      "loss": 3.7556,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291678978067472e-05,
+      "loss": 3.769,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29084038331642e-05,
+      "loss": 3.7542,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290001788565368e-05,
+      "loss": 3.752,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289164831694689e-05,
+      "loss": 3.7625,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28832787482401e-05,
+      "loss": 3.7573,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2874892800729576e-05,
+      "loss": 3.7524,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866506853219056e-05,
+      "loss": 3.755,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2858120905708536e-05,
+      "loss": 3.7541,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849734958198016e-05,
+      "loss": 3.7479,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2841365389491225e-05,
+      "loss": 3.7392,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283297944198071e-05,
+      "loss": 3.7576,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282459349447019e-05,
+      "loss": 3.7543,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281620754695967e-05,
+      "loss": 3.7542,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280783797825288e-05,
+      "loss": 3.7507,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279945203074236e-05,
+      "loss": 3.7458,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279106608323184e-05,
+      "loss": 3.7668,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278268013572132e-05,
+      "loss": 3.7648,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27742941882108e-05,
+      "loss": 3.7611,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276590824070028e-05,
+      "loss": 3.7574,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275752229318976e-05,
+      "loss": 3.7498,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274915272448297e-05,
+      "loss": 3.7498,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274076677697245e-05,
+      "loss": 3.7433,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273238082946193e-05,
+      "loss": 3.7503,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272399488195141e-05,
+      "loss": 3.7525,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715608934440896e-05,
+      "loss": 3.7533,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2707222986930376e-05,
+      "loss": 3.7529,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698853418223585e-05,
+      "loss": 3.7516,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690467470713065e-05,
+      "loss": 3.7539,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2682081523202545e-05,
+      "loss": 3.749,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2673695575692025e-05,
+      "loss": 3.7507,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2665309628181505e-05,
+      "loss": 3.7543,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2656940059474714e-05,
+      "loss": 3.7639,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2648554111964194e-05,
+      "loss": 3.7622,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2640168164453674e-05,
+      "loss": 3.7436,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631782216943154e-05,
+      "loss": 3.7587,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623396269432634e-05,
+      "loss": 3.7504,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261502670072584e-05,
+      "loss": 3.7607,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260664075321533e-05,
+      "loss": 3.7529,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259825480570481e-05,
+      "loss": 3.7539,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258986885819429e-05,
+      "loss": 3.7465,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258148291068376e-05,
+      "loss": 3.7555,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257309696317324e-05,
+      "loss": 3.7425,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256471101566272e-05,
+      "loss": 3.751,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255635782575967e-05,
+      "loss": 3.7494,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254797187824915e-05,
+      "loss": 3.751,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253958593073863e-05,
+      "loss": 3.7531,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253119998322811e-05,
+      "loss": 3.7509,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252281403571758e-05,
+      "loss": 3.7571,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251442808820707e-05,
+      "loss": 3.7423,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250604214069655e-05,
+      "loss": 3.7508,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8681092262268066,
+      "eval_runtime": 311.0138,
+      "eval_samples_per_second": 1226.926,
+      "eval_steps_per_second": 38.342,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249767257198976e-05,
+      "loss": 3.7431,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2489286624479237e-05,
+      "loss": 3.7285,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480900676968716e-05,
+      "loss": 3.7543,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472514729458196e-05,
+      "loss": 3.7484,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464128781947676e-05,
+      "loss": 3.7578,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455742834437156e-05,
+      "loss": 3.7418,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447373265730365e-05,
+      "loss": 3.7388,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2438987318219845e-05,
+      "loss": 3.741,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430601370709325e-05,
+      "loss": 3.7435,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2422215423198805e-05,
+      "loss": 3.7402,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413829475688285e-05,
+      "loss": 3.7488,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24054599069815e-05,
+      "loss": 3.7544,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239707395947098e-05,
+      "loss": 3.7293,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238868801196046e-05,
+      "loss": 3.7374,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238030206444994e-05,
+      "loss": 3.7448,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237193249574315e-05,
+      "loss": 3.7298,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236354654823263e-05,
+      "loss": 3.7379,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235516060072211e-05,
+      "loss": 3.7385,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234677465321159e-05,
+      "loss": 3.7391,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23384050845048e-05,
+      "loss": 3.7551,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233001913699428e-05,
+      "loss": 3.7436,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232163318948376e-05,
+      "loss": 3.7418,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231324724197324e-05,
+      "loss": 3.7417,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304877673266455e-05,
+      "loss": 3.7432,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296491725755935e-05,
+      "loss": 3.7341,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2288105778245415e-05,
+      "loss": 3.7425,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279719830734895e-05,
+      "loss": 3.7338,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2271333883224375e-05,
+      "loss": 3.7313,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262964314517584e-05,
+      "loss": 3.7316,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254578367007064e-05,
+      "loss": 3.736,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246192419496544e-05,
+      "loss": 3.7373,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237806471986024e-05,
+      "loss": 3.7401,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222943690327923e-05,
+      "loss": 3.744,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222105095576871e-05,
+      "loss": 3.7443,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221266500825819e-05,
+      "loss": 3.7324,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220427906074767e-05,
+      "loss": 3.738,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219589311323716e-05,
+      "loss": 3.7362,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218750716572664e-05,
+      "loss": 3.7302,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217913759701985e-05,
+      "loss": 3.7213,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217075164950933e-05,
+      "loss": 3.729,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216236570199881e-05,
+      "loss": 3.7313,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215397975448829e-05,
+      "loss": 3.7375,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214559380697777e-05,
+      "loss": 3.7276,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213722423827098e-05,
+      "loss": 3.7312,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212883829076046e-05,
+      "loss": 3.7494,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212045234324994e-05,
+      "loss": 3.734,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211206639573942e-05,
+      "loss": 3.7398,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103696827032626e-05,
+      "loss": 3.7245,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209531087952211e-05,
+      "loss": 3.7149,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208692493201159e-05,
+      "loss": 3.74,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207853898450107e-05,
+      "loss": 3.7309,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070153036990546e-05,
+      "loss": 3.7265,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206178346828376e-05,
+      "loss": 3.7264,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205339752077324e-05,
+      "loss": 3.7319,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204501157326272e-05,
+      "loss": 3.7114,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036625625752195e-05,
+      "loss": 3.7311,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028239678241675e-05,
+      "loss": 3.7261,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201987010953489e-05,
+      "loss": 3.7206,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011484162024364e-05,
+      "loss": 3.7389,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200309821451385e-05,
+      "loss": 3.7236,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199471226700333e-05,
+      "loss": 3.7228,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1986342698296547e-05,
+      "loss": 3.7309,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197795675078602e-05,
+      "loss": 3.719,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19695708032755e-05,
+      "loss": 3.7191,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196118485576498e-05,
+      "loss": 3.7266,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195281528705819e-05,
+      "loss": 3.7236,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194442933954767e-05,
+      "loss": 3.7242,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193604339203715e-05,
+      "loss": 3.718,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192765744452663e-05,
+      "loss": 3.7087,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191927149701611e-05,
+      "loss": 3.72,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191088554950559e-05,
+      "loss": 3.7354,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902515980798804e-05,
+      "loss": 3.7188,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1894130033288284e-05,
+      "loss": 3.741,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885744085777764e-05,
+      "loss": 3.7331,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1877358138267244e-05,
+      "loss": 3.7314,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186898856956045e-05,
+      "loss": 3.7158,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186060262204993e-05,
+      "loss": 3.7346,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185221667453941e-05,
+      "loss": 3.7224,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184383072702889e-05,
+      "loss": 3.716,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183544477951837e-05,
+      "loss": 3.7341,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182707521081158e-05,
+      "loss": 3.7264,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181868926330106e-05,
+      "loss": 3.731,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181030331579054e-05,
+      "loss": 3.7258,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180191736828002e-05,
+      "loss": 3.7084,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17935314207695e-05,
+      "loss": 3.7082,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178516185206272e-05,
+      "loss": 3.7182,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17767759045522e-05,
+      "loss": 3.7271,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176838995704168e-05,
+      "loss": 3.7199,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176000400953116e-05,
+      "loss": 3.7168,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175163444082437e-05,
+      "loss": 3.7222,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174324849331385e-05,
+      "loss": 3.713,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173486254580333e-05,
+      "loss": 3.7196,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172647659829281e-05,
+      "loss": 3.7146,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1718107029586016e-05,
+      "loss": 3.7342,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709721082075496e-05,
+      "loss": 3.7327,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701335134564976e-05,
+      "loss": 3.7235,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692949187054456e-05,
+      "loss": 3.7149,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168457961834767e-05,
+      "loss": 3.7209,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167619367083715e-05,
+      "loss": 3.7145,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166780772332663e-05,
+      "loss": 3.7326,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165942177581611e-05,
+      "loss": 3.7171,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165105220710932e-05,
+      "loss": 3.7165,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16426662595988e-05,
+      "loss": 3.7207,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163428031208828e-05,
+      "loss": 3.7232,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162589436457776e-05,
+      "loss": 3.7094,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161750841706724e-05,
+      "loss": 3.72,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160912246955672e-05,
+      "loss": 3.7147,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16007365220462e-05,
+      "loss": 3.7102,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159236695333941e-05,
+      "loss": 3.7046,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158398100582889e-05,
+      "loss": 3.7197,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575595058318376e-05,
+      "loss": 3.7205,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1567209110807856e-05,
+      "loss": 3.7127,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558839542101065e-05,
+      "loss": 3.7161,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1550453594590545e-05,
+      "loss": 3.7109,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1542067647080025e-05,
+      "loss": 3.7265,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15336816995695e-05,
+      "loss": 3.7313,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152529575205898e-05,
+      "loss": 3.7257,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151690980454846e-05,
+      "loss": 3.7186,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508540235841674e-05,
+      "loss": 3.7128,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150015428833115e-05,
+      "loss": 3.7151,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149176834082063e-05,
+      "loss": 3.7037,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1483382393310114e-05,
+      "loss": 3.7153,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147501282460333e-05,
+      "loss": 3.716,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14666268770928e-05,
+      "loss": 3.7142,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145824092958228e-05,
+      "loss": 3.7186,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144985498207176e-05,
+      "loss": 3.7157,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144146903456124e-05,
+      "loss": 3.722,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143308308705072e-05,
+      "loss": 3.71,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142471351834393e-05,
+      "loss": 3.7174,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141632757083341e-05,
+      "loss": 3.7181,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140794162332289e-05,
+      "loss": 3.7225,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139955567581237e-05,
+      "loss": 3.7276,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139118610710558e-05,
+      "loss": 3.7077,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138280015959507e-05,
+      "loss": 3.7235,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137441421208455e-05,
+      "loss": 3.7153,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136602826457403e-05,
+      "loss": 3.7256,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357658695867237e-05,
+      "loss": 3.715,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349272748356717e-05,
+      "loss": 3.7204,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340886800846197e-05,
+      "loss": 3.7094,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332500853335676e-05,
+      "loss": 3.7234,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1324131284628886e-05,
+      "loss": 3.7053,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315745337118366e-05,
+      "loss": 3.7163,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307359389607845e-05,
+      "loss": 3.7165,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298973442097325e-05,
+      "loss": 3.7167,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290587494586805e-05,
+      "loss": 3.7147,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128221792588002e-05,
+      "loss": 3.7148,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12738319783695e-05,
+      "loss": 3.7209,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126544603085898e-05,
+      "loss": 3.7063,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125706008334846e-05,
+      "loss": 3.7209,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8568427562713623,
+      "eval_runtime": 304.0833,
+      "eval_samples_per_second": 1254.89,
+      "eval_steps_per_second": 39.216,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124867413583794e-05,
+      "loss": 3.7086,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124030456713115e-05,
+      "loss": 3.6961,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123191861962063e-05,
+      "loss": 3.7159,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122353267211011e-05,
+      "loss": 3.7152,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121514672459959e-05,
+      "loss": 3.7205,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12067771558928e-05,
+      "loss": 3.7066,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119839120838228e-05,
+      "loss": 3.7066,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119000526087176e-05,
+      "loss": 3.7072,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118161931336124e-05,
+      "loss": 3.7101,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1173249744654455e-05,
+      "loss": 3.7037,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1164863797143935e-05,
+      "loss": 3.7143,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156477849633415e-05,
+      "loss": 3.7163,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1148091902122895e-05,
+      "loss": 3.7035,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1139722333416104e-05,
+      "loss": 3.6985,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1131336385905584e-05,
+      "loss": 3.7095,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122950438395064e-05,
+      "loss": 3.691,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114564490884544e-05,
+      "loss": 3.7061,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110619492217775e-05,
+      "loss": 3.7035,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109780897466723e-05,
+      "loss": 3.7078,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108942302715671e-05,
+      "loss": 3.716,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108103707964619e-05,
+      "loss": 3.7125,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107266751093941e-05,
+      "loss": 3.7071,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106428156342889e-05,
+      "loss": 3.7067,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105589561591837e-05,
+      "loss": 3.7062,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104750966840785e-05,
+      "loss": 3.6996,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103914009970106e-05,
+      "loss": 3.7106,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103075415219054e-05,
+      "loss": 3.7012,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102236820468002e-05,
+      "loss": 3.696,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10139822571695e-05,
+      "loss": 3.6951,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100559630965898e-05,
+      "loss": 3.7066,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997226740952186e-05,
+      "loss": 3.6999,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988840793441666e-05,
+      "loss": 3.7093,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980454845931146e-05,
+      "loss": 3.7143,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972068898420626e-05,
+      "loss": 3.7078,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096369932971384e-05,
+      "loss": 3.6957,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095531338220332e-05,
+      "loss": 3.7065,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09469274346928e-05,
+      "loss": 3.7049,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093854148718228e-05,
+      "loss": 3.696,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0930155539671755e-05,
+      "loss": 3.6901,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092178597096497e-05,
+      "loss": 3.692,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091340002345445e-05,
+      "loss": 3.6985,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090501407594393e-05,
+      "loss": 3.7043,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896628128433404e-05,
+      "loss": 3.6933,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088825855972662e-05,
+      "loss": 3.6965,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08798726122161e-05,
+      "loss": 3.7131,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087148666470558e-05,
+      "loss": 3.6992,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086310071719506e-05,
+      "loss": 3.7079,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854731148488276e-05,
+      "loss": 3.693,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0846345200977756e-05,
+      "loss": 3.68,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083795925346723e-05,
+      "loss": 3.7102,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082957330595671e-05,
+      "loss": 3.6989,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082118735844619e-05,
+      "loss": 3.6942,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0812817789739405e-05,
+      "loss": 3.6983,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080443184222888e-05,
+      "loss": 3.694,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079604589471836e-05,
+      "loss": 3.6851,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078765994720784e-05,
+      "loss": 3.6917,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0779290378501054e-05,
+      "loss": 3.6915,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770904430990534e-05,
+      "loss": 3.6921,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762518483480014e-05,
+      "loss": 3.7015,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0754132535969494e-05,
+      "loss": 3.6919,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07457629672627e-05,
+      "loss": 3.692,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073737701975218e-05,
+      "loss": 3.701,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072899107224166e-05,
+      "loss": 3.683,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072060512473114e-05,
+      "loss": 3.6888,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071223555602435e-05,
+      "loss": 3.6912,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070384960851383e-05,
+      "loss": 3.6945,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069546366100331e-05,
+      "loss": 3.6883,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068707771349279e-05,
+      "loss": 3.6904,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067870814478601e-05,
+      "loss": 3.6734,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067032219727549e-05,
+      "loss": 3.688,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066193624976497e-05,
+      "loss": 3.704,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065355030225445e-05,
+      "loss": 3.6851,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0645180733547656e-05,
+      "loss": 3.7082,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636794786037136e-05,
+      "loss": 3.6988,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0628408838526616e-05,
+      "loss": 3.7022,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0620022891016096e-05,
+      "loss": 3.6859,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611653322309305e-05,
+      "loss": 3.6992,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0603267374798785e-05,
+      "loss": 3.691,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594881427288265e-05,
+      "loss": 3.6854,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0586495479777745e-05,
+      "loss": 3.7033,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057812591107096e-05,
+      "loss": 3.6919,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056973996356044e-05,
+      "loss": 3.6985,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056135401604992e-05,
+      "loss": 3.6942,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05529680685394e-05,
+      "loss": 3.672,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054458212102888e-05,
+      "loss": 3.6821,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053621255232209e-05,
+      "loss": 3.6801,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052782660481157e-05,
+      "loss": 3.6971,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051944065730105e-05,
+      "loss": 3.6899,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051105470979053e-05,
+      "loss": 3.6863,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050268514108374e-05,
+      "loss": 3.6874,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049429919357322e-05,
+      "loss": 3.6852,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04859132460627e-05,
+      "loss": 3.687,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047752729855218e-05,
+      "loss": 3.6846,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0469157729845395e-05,
+      "loss": 3.6992,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460771782334875e-05,
+      "loss": 3.7073,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0452385834824355e-05,
+      "loss": 3.6868,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443999887313835e-05,
+      "loss": 3.6846,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435630318607044e-05,
+      "loss": 3.6889,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427244371096524e-05,
+      "loss": 3.6844,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418858423586004e-05,
+      "loss": 3.6995,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0410472476075483e-05,
+      "loss": 3.6861,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040210290736869e-05,
+      "loss": 3.6875,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039371695985817e-05,
+      "loss": 3.6892,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038533101234765e-05,
+      "loss": 3.6919,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037694506483713e-05,
+      "loss": 3.6774,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036855911732661e-05,
+      "loss": 3.6868,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036018954861983e-05,
+      "loss": 3.6861,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035180360110931e-05,
+      "loss": 3.676,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034341765359879e-05,
+      "loss": 3.675,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033503170608827e-05,
+      "loss": 3.6876,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032666213738148e-05,
+      "loss": 3.6876,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031827618987096e-05,
+      "loss": 3.6816,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030989024236044e-05,
+      "loss": 3.6867,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030150429484992e-05,
+      "loss": 3.6807,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0293134726143126e-05,
+      "loss": 3.6976,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284748778632606e-05,
+      "loss": 3.6994,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0276362831122086e-05,
+      "loss": 3.6943,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267976883611566e-05,
+      "loss": 3.6875,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025960731490478e-05,
+      "loss": 3.6786,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025122136739426e-05,
+      "loss": 3.6879,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024283541988374e-05,
+      "loss": 3.6749,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023444947237322e-05,
+      "loss": 3.6845,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022607990366643e-05,
+      "loss": 3.6823,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021769395615591e-05,
+      "loss": 3.6864,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020930800864539e-05,
+      "loss": 3.6841,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020092206113487e-05,
+      "loss": 3.6852,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019255249242808e-05,
+      "loss": 3.6916,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018416654491756e-05,
+      "loss": 3.681,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017578059740704e-05,
+      "loss": 3.6836,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016739464989652e-05,
+      "loss": 3.6881,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0159025081189736e-05,
+      "loss": 3.6879,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0150639133679216e-05,
+      "loss": 3.6985,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0142253186168696e-05,
+      "loss": 3.6789,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0133867238658175e-05,
+      "loss": 3.6946,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0125481291147655e-05,
+      "loss": 3.6847,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0117111722440865e-05,
+      "loss": 3.6924,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108725774930344e-05,
+      "loss": 3.6848,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0100339827419824e-05,
+      "loss": 3.6877,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0091953879909304e-05,
+      "loss": 3.6801,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0083584311202513e-05,
+      "loss": 3.6921,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0075198363691993e-05,
+      "loss": 3.6725,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006681241618147e-05,
+      "loss": 3.6849,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005842646867095e-05,
+      "loss": 3.6879,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005005689996417e-05,
+      "loss": 3.6877,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004167095245365e-05,
+      "loss": 3.6812,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003328500494313e-05,
+      "loss": 3.6832,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00248990574326e-05,
+      "loss": 3.6939,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001652948872582e-05,
+      "loss": 3.6746,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00081435412153e-05,
+      "loss": 3.6915,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.850482225418091,
+      "eval_runtime": 304.9891,
+      "eval_samples_per_second": 1251.163,
+      "eval_steps_per_second": 39.1,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999975759370478e-05,
+      "loss": 3.6696,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999137164619425e-05,
+      "loss": 3.673,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998298569868373e-05,
+      "loss": 3.6813,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997459975117321e-05,
+      "loss": 3.6841,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996623018246643e-05,
+      "loss": 3.6931,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995784423495591e-05,
+      "loss": 3.6752,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994945828744539e-05,
+      "loss": 3.6808,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994107233993487e-05,
+      "loss": 3.6775,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993268639242435e-05,
+      "loss": 3.6753,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9924316823717556e-05,
+      "loss": 3.675,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9915930876207036e-05,
+      "loss": 3.6841,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9907544928696516e-05,
+      "loss": 3.6852,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9899158981185996e-05,
+      "loss": 3.6716,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890789412479205e-05,
+      "loss": 3.6693,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882403464968685e-05,
+      "loss": 3.6831,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9874017517458165e-05,
+      "loss": 3.6603,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865631569947645e-05,
+      "loss": 3.6756,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985726200124086e-05,
+      "loss": 3.6731,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984887605373034e-05,
+      "loss": 3.6804,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984049010621982e-05,
+      "loss": 3.6805,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98321041587093e-05,
+      "loss": 3.6866,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982373459000251e-05,
+      "loss": 3.6783,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981534864249199e-05,
+      "loss": 3.6781,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980696269498147e-05,
+      "loss": 3.6727,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979857674747095e-05,
+      "loss": 3.6719,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979020717876416e-05,
+      "loss": 3.6827,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978182123125364e-05,
+      "loss": 3.6724,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977343528374312e-05,
+      "loss": 3.667,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97650493362326e-05,
+      "loss": 3.6718,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9756679767525814e-05,
+      "loss": 3.6738,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9748293820015294e-05,
+      "loss": 3.6693,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739907872504774e-05,
+      "loss": 3.6832,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9731521924994254e-05,
+      "loss": 3.6806,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972315235628746e-05,
+      "loss": 3.6763,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971476640877694e-05,
+      "loss": 3.6679,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970638046126642e-05,
+      "loss": 3.6819,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96979945137559e-05,
+      "loss": 3.6747,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968962494504911e-05,
+      "loss": 3.6683,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968123899753859e-05,
+      "loss": 3.66,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967285305002807e-05,
+      "loss": 3.6638,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966446710251755e-05,
+      "loss": 3.6665,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965609753381077e-05,
+      "loss": 3.6793,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964771158630025e-05,
+      "loss": 3.6673,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963932563878973e-05,
+      "loss": 3.6657,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963093969127921e-05,
+      "loss": 3.6828,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962257012257242e-05,
+      "loss": 3.6707,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96141841750619e-05,
+      "loss": 3.6815,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960579822755138e-05,
+      "loss": 3.6652,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959741228004086e-05,
+      "loss": 3.6476,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9589042711334066e-05,
+      "loss": 3.6816,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9580656763823546e-05,
+      "loss": 3.6726,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9572270816313026e-05,
+      "loss": 3.666,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9563884868802506e-05,
+      "loss": 3.6666,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955551530009572e-05,
+      "loss": 3.6668,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95471293525852e-05,
+      "loss": 3.6552,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953874340507468e-05,
+      "loss": 3.6608,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953035745756416e-05,
+      "loss": 3.6642,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952197151005364e-05,
+      "loss": 3.6639,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951360194134685e-05,
+      "loss": 3.6767,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950521599383633e-05,
+      "loss": 3.6627,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949683004632581e-05,
+      "loss": 3.6619,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948844409881529e-05,
+      "loss": 3.6779,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94800745301085e-05,
+      "loss": 3.6517,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947168858259798e-05,
+      "loss": 3.6582,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946330263508746e-05,
+      "loss": 3.662,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945491668757694e-05,
+      "loss": 3.6647,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9446547118870155e-05,
+      "loss": 3.6606,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9438161171359635e-05,
+      "loss": 3.6649,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429775223849115e-05,
+      "loss": 3.6468,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9421389276338595e-05,
+      "loss": 3.6584,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9413019707631804e-05,
+      "loss": 3.673,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9404633760121284e-05,
+      "loss": 3.6581,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9396247812610764e-05,
+      "loss": 3.6797,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387861865100244e-05,
+      "loss": 3.671,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937949229639345e-05,
+      "loss": 3.6775,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937110634888293e-05,
+      "loss": 3.6546,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936272040137241e-05,
+      "loss": 3.6699,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935433445386189e-05,
+      "loss": 3.6667,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934596488515511e-05,
+      "loss": 3.6536,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933757893764459e-05,
+      "loss": 3.6743,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932919299013407e-05,
+      "loss": 3.666,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932080704262355e-05,
+      "loss": 3.6673,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931243747391676e-05,
+      "loss": 3.6669,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930405152640624e-05,
+      "loss": 3.6486,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929566557889572e-05,
+      "loss": 3.6527,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92872796313852e-05,
+      "loss": 3.6567,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927891006267841e-05,
+      "loss": 3.6658,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927052411516789e-05,
+      "loss": 3.6613,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926213816765737e-05,
+      "loss": 3.6583,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925375222014685e-05,
+      "loss": 3.6617,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924538265144006e-05,
+      "loss": 3.6595,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923699670392954e-05,
+      "loss": 3.6581,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922861075641902e-05,
+      "loss": 3.658,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92202248089085e-05,
+      "loss": 3.6752,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211838861397976e-05,
+      "loss": 3.6728,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920346929269119e-05,
+      "loss": 3.6651,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919508334518067e-05,
+      "loss": 3.6547,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918669739767015e-05,
+      "loss": 3.6611,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917832782896336e-05,
+      "loss": 3.6561,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916994188145284e-05,
+      "loss": 3.6712,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916155593394232e-05,
+      "loss": 3.6601,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91531699864318e-05,
+      "loss": 3.6582,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144800417725016e-05,
+      "loss": 3.6627,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9136414470214496e-05,
+      "loss": 3.6674,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9128028522703976e-05,
+      "loss": 3.6511,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911964257519345e-05,
+      "loss": 3.6599,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9111273006486665e-05,
+      "loss": 3.6566,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102887058976145e-05,
+      "loss": 3.6545,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9094501111465625e-05,
+      "loss": 3.6485,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90861151639551e-05,
+      "loss": 3.6599,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077745595248314e-05,
+      "loss": 3.658,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9069359647737794e-05,
+      "loss": 3.6608,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060973700227274e-05,
+      "loss": 3.6594,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9052587752716754e-05,
+      "loss": 3.6524,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904421818400997e-05,
+      "loss": 3.6686,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903583223649945e-05,
+      "loss": 3.6728,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902744628898892e-05,
+      "loss": 3.6616,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90190603414784e-05,
+      "loss": 3.6693,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901069077277162e-05,
+      "loss": 3.6469,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90023048252611e-05,
+      "loss": 3.6622,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899391887775057e-05,
+      "loss": 3.654,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898553293024005e-05,
+      "loss": 3.6564,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897716336153327e-05,
+      "loss": 3.652,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896877741402275e-05,
+      "loss": 3.6569,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896039146651223e-05,
+      "loss": 3.6592,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895200551900171e-05,
+      "loss": 3.6573,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8943635950294924e-05,
+      "loss": 3.6664,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89352500027844e-05,
+      "loss": 3.6553,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892686405527388e-05,
+      "loss": 3.6533,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891847810776336e-05,
+      "loss": 3.6634,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891010853905657e-05,
+      "loss": 3.6636,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8901722591546046e-05,
+      "loss": 3.6722,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8893336644035526e-05,
+      "loss": 3.6509,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884950696525006e-05,
+      "loss": 3.6683,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8876564749014486e-05,
+      "loss": 3.6578,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88681951803077e-05,
+      "loss": 3.6654,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885980923279718e-05,
+      "loss": 3.6599,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885142328528666e-05,
+      "loss": 3.6615,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884303733777614e-05,
+      "loss": 3.6525,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883466776906935e-05,
+      "loss": 3.6685,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882628182155883e-05,
+      "loss": 3.6489,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881789587404831e-05,
+      "loss": 3.6595,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880950992653779e-05,
+      "loss": 3.6606,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8801140357831e-05,
+      "loss": 3.665,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879275441032048e-05,
+      "loss": 3.6527,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878436846280996e-05,
+      "loss": 3.6584,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877598251529944e-05,
+      "loss": 3.6641,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8767612946592655e-05,
+      "loss": 3.6553,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8759226999082135e-05,
+      "loss": 3.6633,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8461461067199707,
+      "eval_runtime": 303.6905,
+      "eval_samples_per_second": 1256.513,
+      "eval_steps_per_second": 39.267,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750841051571615e-05,
+      "loss": 3.6446,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8742455104061095e-05,
+      "loss": 3.6496,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8734085535354304e-05,
+      "loss": 3.6534,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8725699587843784e-05,
+      "loss": 3.6593,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8717313640333264e-05,
+      "loss": 3.667,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8708927692822744e-05,
+      "loss": 3.6489,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870055812411595e-05,
+      "loss": 3.6547,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869217217660543e-05,
+      "loss": 3.6539,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868378622909491e-05,
+      "loss": 3.6462,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867540028158439e-05,
+      "loss": 3.6514,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866703071287761e-05,
+      "loss": 3.6595,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865864476536709e-05,
+      "loss": 3.6564,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865025881785657e-05,
+      "loss": 3.6565,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864187287034605e-05,
+      "loss": 3.6358,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863350330163926e-05,
+      "loss": 3.6568,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862511735412874e-05,
+      "loss": 3.6355,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861673140661822e-05,
+      "loss": 3.6481,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86083454591077e-05,
+      "loss": 3.6517,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859997589040091e-05,
+      "loss": 3.6509,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859158994289039e-05,
+      "loss": 3.6552,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858320399537987e-05,
+      "loss": 3.6609,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857481804786935e-05,
+      "loss": 3.6576,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8566448479162556e-05,
+      "loss": 3.6541,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855806253165204e-05,
+      "loss": 3.6432,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854967658414152e-05,
+      "loss": 3.649,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8541290636631e-05,
+      "loss": 3.6537,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853292106792421e-05,
+      "loss": 3.6506,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852453512041369e-05,
+      "loss": 3.6448,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851614917290317e-05,
+      "loss": 3.6415,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850776322539265e-05,
+      "loss": 3.6505,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849939365668586e-05,
+      "loss": 3.6409,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849100770917534e-05,
+      "loss": 3.6589,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848262176166482e-05,
+      "loss": 3.6538,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84742358141543e-05,
+      "loss": 3.6532,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846586624544751e-05,
+      "loss": 3.6441,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8457480297936996e-05,
+      "loss": 3.6507,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8449094350426476e-05,
+      "loss": 3.6534,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8440708402915956e-05,
+      "loss": 3.6439,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8432338834209165e-05,
+      "loss": 3.6383,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423952886698645e-05,
+      "loss": 3.6391,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8415566939188125e-05,
+      "loss": 3.6398,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8407180991677605e-05,
+      "loss": 3.654,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8398811422970814e-05,
+      "loss": 3.6415,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8390425475460294e-05,
+      "loss": 3.6436,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8382039527949774e-05,
+      "loss": 3.6562,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837366995924298e-05,
+      "loss": 3.6468,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836528401173246e-05,
+      "loss": 3.654,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835689806422195e-05,
+      "loss": 3.6421,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834851211671143e-05,
+      "loss": 3.6252,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834014254800464e-05,
+      "loss": 3.6561,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833175660049412e-05,
+      "loss": 3.6439,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83233706529836e-05,
+      "loss": 3.6454,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831498470547308e-05,
+      "loss": 3.6436,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830659875796256e-05,
+      "loss": 3.6425,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829822918925577e-05,
+      "loss": 3.6263,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828984324174525e-05,
+      "loss": 3.6392,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828145729423473e-05,
+      "loss": 3.6367,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827307134672421e-05,
+      "loss": 3.6358,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826470177801742e-05,
+      "loss": 3.6547,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8256315830506903e-05,
+      "loss": 3.6372,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8247929882996383e-05,
+      "loss": 3.6361,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823954393548586e-05,
+      "loss": 3.6539,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823117436677907e-05,
+      "loss": 3.6277,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822278841926855e-05,
+      "loss": 3.6307,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821440247175803e-05,
+      "loss": 3.6424,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820601652424751e-05,
+      "loss": 3.6413,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819764695554072e-05,
+      "loss": 3.6331,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81892610080302e-05,
+      "loss": 3.6432,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818087506051968e-05,
+      "loss": 3.6236,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817250549181289e-05,
+      "loss": 3.6335,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816411954430237e-05,
+      "loss": 3.6497,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815573359679186e-05,
+      "loss": 3.6367,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814734764928134e-05,
+      "loss": 3.6503,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8138978080574546e-05,
+      "loss": 3.6457,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8130592133064026e-05,
+      "loss": 3.657,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8122206185553506e-05,
+      "loss": 3.6313,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8113820238042986e-05,
+      "loss": 3.6452,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8105450669336195e-05,
+      "loss": 3.6443,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8097064721825675e-05,
+      "loss": 3.6278,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8088678774315155e-05,
+      "loss": 3.6532,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8080292826804635e-05,
+      "loss": 3.6395,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071923258097844e-05,
+      "loss": 3.6435,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8063537310587324e-05,
+      "loss": 3.644,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805515136307681e-05,
+      "loss": 3.6268,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804676541556629e-05,
+      "loss": 3.625,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80383958468595e-05,
+      "loss": 3.6331,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803000989934898e-05,
+      "loss": 3.6413,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802162395183846e-05,
+      "loss": 3.6376,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801323800432794e-05,
+      "loss": 3.6333,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800486843562115e-05,
+      "loss": 3.6344,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799648248811063e-05,
+      "loss": 3.641,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798809654060011e-05,
+      "loss": 3.6339,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797971059308959e-05,
+      "loss": 3.629,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79713410243828e-05,
+      "loss": 3.6521,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796295507687228e-05,
+      "loss": 3.6505,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7954569129361764e-05,
+      "loss": 3.6443,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7946183181851244e-05,
+      "loss": 3.6282,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793779723434072e-05,
+      "loss": 3.6389,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7929427665633933e-05,
+      "loss": 3.6306,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7921041718123413e-05,
+      "loss": 3.645,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7912655770612893e-05,
+      "loss": 3.6374,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7904269823102367e-05,
+      "loss": 3.6328,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789590025439558e-05,
+      "loss": 3.638,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788751430688506e-05,
+      "loss": 3.6463,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787912835937454e-05,
+      "loss": 3.628,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7870742411864016e-05,
+      "loss": 3.6385,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786237284315723e-05,
+      "loss": 3.6282,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785398689564671e-05,
+      "loss": 3.6334,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784560094813619e-05,
+      "loss": 3.6266,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783721500062567e-05,
+      "loss": 3.6302,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782884543191889e-05,
+      "loss": 3.6346,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782045948440837e-05,
+      "loss": 3.6385,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781207353689784e-05,
+      "loss": 3.6357,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780368758938732e-05,
+      "loss": 3.6269,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7795318020680536e-05,
+      "loss": 3.6457,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7786932073170016e-05,
+      "loss": 3.6476,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777854612565949e-05,
+      "loss": 3.6389,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777016017814897e-05,
+      "loss": 3.6484,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7761790609442185e-05,
+      "loss": 3.6239,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7753404661931665e-05,
+      "loss": 3.6343,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7745018714421145e-05,
+      "loss": 3.6317,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7736632766910625e-05,
+      "loss": 3.6277,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772826319820384e-05,
+      "loss": 3.6307,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7719877250693314e-05,
+      "loss": 3.6335,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7711491303182794e-05,
+      "loss": 3.6353,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7703105355672274e-05,
+      "loss": 3.633,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769473578696549e-05,
+      "loss": 3.6404,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768634983945496e-05,
+      "loss": 3.6363,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767796389194444e-05,
+      "loss": 3.6284,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766957794443392e-05,
+      "loss": 3.6375,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766120837572714e-05,
+      "loss": 3.6438,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765282242821662e-05,
+      "loss": 3.6458,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76444364807061e-05,
+      "loss": 3.6289,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763605053319558e-05,
+      "loss": 3.648,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762768096448879e-05,
+      "loss": 3.633,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761929501697827e-05,
+      "loss": 3.6413,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761090906946775e-05,
+      "loss": 3.6378,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760252312195723e-05,
+      "loss": 3.6355,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759415355325044e-05,
+      "loss": 3.6294,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758576760573992e-05,
+      "loss": 3.6481,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7577381658229397e-05,
+      "loss": 3.6183,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7568995710718877e-05,
+      "loss": 3.6395,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756062614201209e-05,
+      "loss": 3.6407,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755224019450157e-05,
+      "loss": 3.6354,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754385424699105e-05,
+      "loss": 3.6314,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753546829948053e-05,
+      "loss": 3.6344,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752709873077374e-05,
+      "loss": 3.6382,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751871278326322e-05,
+      "loss": 3.6363,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75103268357527e-05,
+      "loss": 3.637,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.843632698059082,
+      "eval_runtime": 303.9931,
+      "eval_samples_per_second": 1255.262,
+      "eval_steps_per_second": 39.228,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750194088824218e-05,
+      "loss": 3.6133,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749355494073166e-05,
+      "loss": 3.6266,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748516899322114e-05,
+      "loss": 3.6327,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747678304571062e-05,
+      "loss": 3.6361,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746841347700383e-05,
+      "loss": 3.6443,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746002752949331e-05,
+      "loss": 3.6254,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74516415819828e-05,
+      "loss": 3.6333,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744325563447228e-05,
+      "loss": 3.6297,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7434886065765486e-05,
+      "loss": 3.6254,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7426500118254966e-05,
+      "loss": 3.6257,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7418114170744446e-05,
+      "loss": 3.6338,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7409728223233926e-05,
+      "loss": 3.6356,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7401358654527135e-05,
+      "loss": 3.6364,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7392972707016615e-05,
+      "loss": 3.6116,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7384586759506095e-05,
+      "loss": 3.6348,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7376200811995575e-05,
+      "loss": 3.6108,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7367831243288784e-05,
+      "loss": 3.625,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7359445295778264e-05,
+      "loss": 3.6253,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735105934826775e-05,
+      "loss": 3.6278,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734267340075723e-05,
+      "loss": 3.6335,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733430383205044e-05,
+      "loss": 3.639,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732591788453992e-05,
+      "loss": 3.6336,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73175319370294e-05,
+      "loss": 3.6281,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730914598951888e-05,
+      "loss": 3.6192,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730077642081209e-05,
+      "loss": 3.6302,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729239047330157e-05,
+      "loss": 3.6321,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728400452579105e-05,
+      "loss": 3.6255,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727561857828053e-05,
+      "loss": 3.6228,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726724900957374e-05,
+      "loss": 3.6172,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725886306206322e-05,
+      "loss": 3.6288,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7250477114552704e-05,
+      "loss": 3.6158,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7242091167042184e-05,
+      "loss": 3.6356,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723372159833539e-05,
+      "loss": 3.6384,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722533565082487e-05,
+      "loss": 3.6269,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721694970331435e-05,
+      "loss": 3.6251,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720856375580383e-05,
+      "loss": 3.6224,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720019418709704e-05,
+      "loss": 3.6339,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719180823958652e-05,
+      "loss": 3.6226,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7183422292076e-05,
+      "loss": 3.6099,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717503634456548e-05,
+      "loss": 3.6189,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716666677585869e-05,
+      "loss": 3.6181,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715828082834817e-05,
+      "loss": 3.6282,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714989488083766e-05,
+      "loss": 3.622,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714150893332714e-05,
+      "loss": 3.621,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713313936462035e-05,
+      "loss": 3.63,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712475341710983e-05,
+      "loss": 3.626,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711636746959931e-05,
+      "loss": 3.6305,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710798152208879e-05,
+      "loss": 3.6245,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7099611953381996e-05,
+      "loss": 3.598,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7091226005871476e-05,
+      "loss": 3.6342,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7082840058360956e-05,
+      "loss": 3.6229,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7074454110850436e-05,
+      "loss": 3.6215,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7066084542143645e-05,
+      "loss": 3.6254,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7057698594633125e-05,
+      "loss": 3.6215,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704931264712261e-05,
+      "loss": 3.6052,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704092669961209e-05,
+      "loss": 3.6145,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70325571309053e-05,
+      "loss": 3.6193,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702417118339478e-05,
+      "loss": 3.6145,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701578523588426e-05,
+      "loss": 3.6297,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700739928837374e-05,
+      "loss": 3.616,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699902971966695e-05,
+      "loss": 3.6142,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699064377215643e-05,
+      "loss": 3.6301,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698225782464591e-05,
+      "loss": 3.6086,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697387187713539e-05,
+      "loss": 3.608,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69655023084286e-05,
+      "loss": 3.6233,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695711636091808e-05,
+      "loss": 3.6141,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694873041340756e-05,
+      "loss": 3.6123,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6940344465897045e-05,
+      "loss": 3.6176,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6931974897190254e-05,
+      "loss": 3.6041,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6923588949679734e-05,
+      "loss": 3.6123,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6915203002169214e-05,
+      "loss": 3.6283,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690681705465869e-05,
+      "loss": 3.6158,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68984474859519e-05,
+      "loss": 3.6277,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689006153844138e-05,
+      "loss": 3.6204,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688167559093086e-05,
+      "loss": 3.6368,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6873289643420336e-05,
+      "loss": 3.6081,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686492007471355e-05,
+      "loss": 3.624,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685653412720303e-05,
+      "loss": 3.6218,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684814817969251e-05,
+      "loss": 3.6053,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683976223218199e-05,
+      "loss": 3.629,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683139266347521e-05,
+      "loss": 3.619,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682300671596469e-05,
+      "loss": 3.6231,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681462076845416e-05,
+      "loss": 3.6238,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680623482094364e-05,
+      "loss": 3.6078,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679786525223686e-05,
+      "loss": 3.6033,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678947930472634e-05,
+      "loss": 3.6073,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678109335721581e-05,
+      "loss": 3.6226,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677270740970529e-05,
+      "loss": 3.6131,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6764337840998506e-05,
+      "loss": 3.6121,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6755951893487986e-05,
+      "loss": 3.6101,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6747565945977466e-05,
+      "loss": 3.6201,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6739179998466946e-05,
+      "loss": 3.6104,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673081042976016e-05,
+      "loss": 3.6133,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6722424482249635e-05,
+      "loss": 3.6266,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6714038534739115e-05,
+      "loss": 3.6297,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6705652587228595e-05,
+      "loss": 3.62,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669728301852181e-05,
+      "loss": 3.6092,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6688897071011284e-05,
+      "loss": 3.6145,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6680511123500764e-05,
+      "loss": 3.6072,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6672125175990244e-05,
+      "loss": 3.6236,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666375560728346e-05,
+      "loss": 3.6148,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665536965977294e-05,
+      "loss": 3.6177,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664698371226242e-05,
+      "loss": 3.6123,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66385977647519e-05,
+      "loss": 3.6191,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663022819604511e-05,
+      "loss": 3.6142,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662184224853459e-05,
+      "loss": 3.6139,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661345630102407e-05,
+      "loss": 3.608,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660507035351355e-05,
+      "loss": 3.6117,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659670078480676e-05,
+      "loss": 3.6066,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658831483729624e-05,
+      "loss": 3.6076,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657992888978572e-05,
+      "loss": 3.6139,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65715429422752e-05,
+      "loss": 3.6127,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656317337356841e-05,
+      "loss": 3.6153,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655478742605789e-05,
+      "loss": 3.6064,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654640147854737e-05,
+      "loss": 3.625,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653801553103685e-05,
+      "loss": 3.6299,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652964596233006e-05,
+      "loss": 3.6167,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652126001481954e-05,
+      "loss": 3.6217,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651287406730902e-05,
+      "loss": 3.6107,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65044881197985e-05,
+      "loss": 3.612,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649611855109171e-05,
+      "loss": 3.6098,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648773260358119e-05,
+      "loss": 3.6056,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647934665607067e-05,
+      "loss": 3.6096,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647096070856015e-05,
+      "loss": 3.6155,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646259113985337e-05,
+      "loss": 3.6176,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645420519234285e-05,
+      "loss": 3.6142,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644581924483233e-05,
+      "loss": 3.6188,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643743329732181e-05,
+      "loss": 3.6105,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6429063728615016e-05,
+      "loss": 3.6101,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6420677781104496e-05,
+      "loss": 3.6134,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6412291833593976e-05,
+      "loss": 3.6243,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6403905886083456e-05,
+      "loss": 3.6237,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6395536317376665e-05,
+      "loss": 3.6067,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6387150369866145e-05,
+      "loss": 3.6273,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6378764422355625e-05,
+      "loss": 3.6128,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6370378474845105e-05,
+      "loss": 3.622,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636200890613832e-05,
+      "loss": 3.6159,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63536229586278e-05,
+      "loss": 3.6093,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634523701111728e-05,
+      "loss": 3.6066,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633685106360676e-05,
+      "loss": 3.6287,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632848149489997e-05,
+      "loss": 3.6013,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632009554738945e-05,
+      "loss": 3.6153,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631170959987893e-05,
+      "loss": 3.6162,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630332365236841e-05,
+      "loss": 3.6176,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629495408366162e-05,
+      "loss": 3.6098,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.62865681361511e-05,
+      "loss": 3.6126,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627818218864058e-05,
+      "loss": 3.6135,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626981261993379e-05,
+      "loss": 3.6226,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6261426672423274e-05,
+      "loss": 3.6162,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8418772220611572,
+      "eval_runtime": 303.6353,
+      "eval_samples_per_second": 1256.741,
+      "eval_steps_per_second": 39.274,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6253040724912754e-05,
+      "loss": 3.5993,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6244654777402234e-05,
+      "loss": 3.6036,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6236268829891714e-05,
+      "loss": 3.6134,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622789926118492e-05,
+      "loss": 3.6149,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62195133136744e-05,
+      "loss": 3.624,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621112736616388e-05,
+      "loss": 3.6078,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620275779745709e-05,
+      "loss": 3.6115,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619437184994657e-05,
+      "loss": 3.6108,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618598590243605e-05,
+      "loss": 3.6045,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617759995492553e-05,
+      "loss": 3.6009,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616923038621874e-05,
+      "loss": 3.6134,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616084443870823e-05,
+      "loss": 3.6207,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615245849119771e-05,
+      "loss": 3.6125,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614407254368719e-05,
+      "loss": 3.5909,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61357029749804e-05,
+      "loss": 3.6129,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612731702746988e-05,
+      "loss": 3.5915,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611893107995936e-05,
+      "loss": 3.6043,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611054513244884e-05,
+      "loss": 3.6079,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6102175563742046e-05,
+      "loss": 3.6075,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6093789616231526e-05,
+      "loss": 3.6087,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6085403668721006e-05,
+      "loss": 3.6179,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6077017721210486e-05,
+      "loss": 3.6115,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6068648152503695e-05,
+      "loss": 3.6118,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606026220499318e-05,
+      "loss": 3.6024,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605187625748266e-05,
+      "loss": 3.6078,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604349030997214e-05,
+      "loss": 3.6099,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603512074126535e-05,
+      "loss": 3.602,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602673479375483e-05,
+      "loss": 3.6077,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601834884624431e-05,
+      "loss": 3.5968,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600996289873379e-05,
+      "loss": 3.6066,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6001593330027e-05,
+      "loss": 3.6016,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599320738251648e-05,
+      "loss": 3.612,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598482143500596e-05,
+      "loss": 3.6117,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597643548749544e-05,
+      "loss": 3.6133,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596806591878865e-05,
+      "loss": 3.6046,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595967997127813e-05,
+      "loss": 3.6039,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5951294023767615e-05,
+      "loss": 3.6145,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5942908076257095e-05,
+      "loss": 3.5974,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5934538507550304e-05,
+      "loss": 3.5949,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5926152560039784e-05,
+      "loss": 3.6015,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5917766612529264e-05,
+      "loss": 3.5972,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5909380665018744e-05,
+      "loss": 3.6031,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590101109631195e-05,
+      "loss": 3.5988,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589262514880143e-05,
+      "loss": 3.6047,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588423920129091e-05,
+      "loss": 3.6052,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587585325378039e-05,
+      "loss": 3.6083,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58674836850736e-05,
+      "loss": 3.6088,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585909773756308e-05,
+      "loss": 3.6047,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585071179005257e-05,
+      "loss": 3.5814,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584232584254205e-05,
+      "loss": 3.6109,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583395627383526e-05,
+      "loss": 3.6037,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582557032632474e-05,
+      "loss": 3.6016,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581718437881422e-05,
+      "loss": 3.6044,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58087984313037e-05,
+      "loss": 3.5985,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580042886259691e-05,
+      "loss": 3.5848,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579204291508639e-05,
+      "loss": 3.5943,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578365696757587e-05,
+      "loss": 3.5997,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577527102006535e-05,
+      "loss": 3.593,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5766901451358556e-05,
+      "loss": 3.6131,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5758515503848036e-05,
+      "loss": 3.5999,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575012955633752e-05,
+      "loss": 3.5872,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5741743608827e-05,
+      "loss": 3.6118,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573337404012021e-05,
+      "loss": 3.5921,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572498809260969e-05,
+      "loss": 3.5867,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571660214509917e-05,
+      "loss": 3.6026,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570821619758865e-05,
+      "loss": 3.5965,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569984662888186e-05,
+      "loss": 3.5921,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569146068137134e-05,
+      "loss": 3.5969,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568307473386082e-05,
+      "loss": 3.5845,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56746887863503e-05,
+      "loss": 3.5942,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566631921764351e-05,
+      "loss": 3.6045,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565793327013299e-05,
+      "loss": 3.5997,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5649547322622476e-05,
+      "loss": 3.6097,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5641161375111956e-05,
+      "loss": 3.5969,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5632791806405165e-05,
+      "loss": 3.6169,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5624405858894645e-05,
+      "loss": 3.5882,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5616019911384125e-05,
+      "loss": 3.6069,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5607633963873605e-05,
+      "loss": 3.5997,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5599264395166814e-05,
+      "loss": 3.5866,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5590878447656294e-05,
+      "loss": 3.6056,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5582492500145774e-05,
+      "loss": 3.6049,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5574106552635254e-05,
+      "loss": 3.5991,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556573698392846e-05,
+      "loss": 3.6055,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555735103641794e-05,
+      "loss": 3.5852,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554896508890743e-05,
+      "loss": 3.5829,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55405791413969e-05,
+      "loss": 3.5894,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553220957269012e-05,
+      "loss": 3.6033,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55238236251796e-05,
+      "loss": 3.5933,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551543767766908e-05,
+      "loss": 3.5956,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550705173015855e-05,
+      "loss": 3.5931,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549868216145177e-05,
+      "loss": 3.5954,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549029621394125e-05,
+      "loss": 3.5927,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548191026643072e-05,
+      "loss": 3.5962,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54735243189202e-05,
+      "loss": 3.6031,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546515475021342e-05,
+      "loss": 3.6097,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54567688027029e-05,
+      "loss": 3.6008,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544838285519238e-05,
+      "loss": 3.591,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543999690768186e-05,
+      "loss": 3.5914,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543162733897507e-05,
+      "loss": 3.5904,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542324139146455e-05,
+      "loss": 3.6023,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5414855443954026e-05,
+      "loss": 3.6003,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5406469496443506e-05,
+      "loss": 3.5971,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539809992773672e-05,
+      "loss": 3.5946,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5389713980226195e-05,
+      "loss": 3.5951,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5381328032715675e-05,
+      "loss": 3.5992,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5372942085205155e-05,
+      "loss": 3.5922,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536457251649837e-05,
+      "loss": 3.5884,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535618656898785e-05,
+      "loss": 3.589,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534780062147733e-05,
+      "loss": 3.5904,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533941467396681e-05,
+      "loss": 3.5877,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5331045105260026e-05,
+      "loss": 3.5935,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53226591577495e-05,
+      "loss": 3.5925,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531427321023898e-05,
+      "loss": 3.595,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530588726272846e-05,
+      "loss": 3.5873,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529751769402167e-05,
+      "loss": 3.6056,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528913174651115e-05,
+      "loss": 3.6098,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528074579900063e-05,
+      "loss": 3.5974,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527235985149011e-05,
+      "loss": 3.6009,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5263990282783324e-05,
+      "loss": 3.5914,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5255604335272804e-05,
+      "loss": 3.5922,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5247218387762284e-05,
+      "loss": 3.5936,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5238832440251764e-05,
+      "loss": 3.5833,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523046287154497e-05,
+      "loss": 3.5934,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522207692403445e-05,
+      "loss": 3.5957,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521369097652393e-05,
+      "loss": 3.6018,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520530502901341e-05,
+      "loss": 3.5904,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519693546030662e-05,
+      "loss": 3.6011,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51885495127961e-05,
+      "loss": 3.5917,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518016356528558e-05,
+      "loss": 3.5901,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517177761777506e-05,
+      "loss": 3.5963,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516340804906828e-05,
+      "loss": 3.6017,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515502210155776e-05,
+      "loss": 3.6051,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514663615404724e-05,
+      "loss": 3.5855,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513825020653672e-05,
+      "loss": 3.6085,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512988063782993e-05,
+      "loss": 3.5924,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512149469031941e-05,
+      "loss": 3.6024,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511310874280889e-05,
+      "loss": 3.5985,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5104739174102096e-05,
+      "loss": 3.5891,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5096353226591576e-05,
+      "loss": 3.591,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5087967279081056e-05,
+      "loss": 3.6099,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5079581331570536e-05,
+      "loss": 3.583,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507121176286375e-05,
+      "loss": 3.5962,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506282581535323e-05,
+      "loss": 3.5958,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505443986784271e-05,
+      "loss": 3.5952,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504605392033219e-05,
+      "loss": 3.5953,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503766797282167e-05,
+      "loss": 3.5917,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502929840411488e-05,
+      "loss": 3.596,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502091245660436e-05,
+      "loss": 3.6043,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501252650909384e-05,
+      "loss": 3.5939,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8409719467163086,
+      "eval_runtime": 304.3782,
+      "eval_samples_per_second": 1253.674,
+      "eval_steps_per_second": 39.178,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500414056158332e-05,
+      "loss": 3.5825,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49957546140728e-05,
+      "loss": 3.5806,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498736866656228e-05,
+      "loss": 3.5929,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497898271905176e-05,
+      "loss": 3.5998,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497061315034497e-05,
+      "loss": 3.597,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496222720283445e-05,
+      "loss": 3.5906,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495384125532393e-05,
+      "loss": 3.5975,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4945455307813416e-05,
+      "loss": 3.5881,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4937085739106625e-05,
+      "loss": 3.5849,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4928699791596105e-05,
+      "loss": 3.5788,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4920313844085585e-05,
+      "loss": 3.5987,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4911927896575065e-05,
+      "loss": 3.5991,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4903558327868274e-05,
+      "loss": 3.5954,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4895172380357754e-05,
+      "loss": 3.5723,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4886786432847234e-05,
+      "loss": 3.5928,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4878400485336714e-05,
+      "loss": 3.5719,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487003091662992e-05,
+      "loss": 3.5894,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48616449691194e-05,
+      "loss": 3.5867,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485325902160888e-05,
+      "loss": 3.5876,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484487307409837e-05,
+      "loss": 3.5918,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483650350539158e-05,
+      "loss": 3.6007,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482811755788106e-05,
+      "loss": 3.5907,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481973161037054e-05,
+      "loss": 3.5904,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481134566286002e-05,
+      "loss": 3.587,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480297609415323e-05,
+      "loss": 3.5897,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479459014664271e-05,
+      "loss": 3.5933,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478622057793592e-05,
+      "loss": 3.5836,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.47778346304254e-05,
+      "loss": 3.585,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4769448682914877e-05,
+      "loss": 3.5797,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4761062735404357e-05,
+      "loss": 3.5864,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4752676787893837e-05,
+      "loss": 3.5858,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474429084038332e-05,
+      "loss": 3.5906,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47359048928728e-05,
+      "loss": 3.5877,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4727518945362276e-05,
+      "loss": 3.5973,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471914937665549e-05,
+      "loss": 3.5859,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471076342914497e-05,
+      "loss": 3.5838,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470237748163445e-05,
+      "loss": 3.5976,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4693991534123925e-05,
+      "loss": 3.5778,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468562196541714e-05,
+      "loss": 3.5782,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467723601790662e-05,
+      "loss": 3.5822,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46688500703961e-05,
+      "loss": 3.5804,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4660464122885574e-05,
+      "loss": 3.5829,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465209455417879e-05,
+      "loss": 3.5822,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464370860666828e-05,
+      "loss": 3.585,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463532265915775e-05,
+      "loss": 3.5858,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462693671164723e-05,
+      "loss": 3.5896,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4618567142940446e-05,
+      "loss": 3.5916,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4610181195429926e-05,
+      "loss": 3.5876,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46017952479194e-05,
+      "loss": 3.5646,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459340930040888e-05,
+      "loss": 3.5905,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4585039731702095e-05,
+      "loss": 3.5827,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4576653784191575e-05,
+      "loss": 3.5842,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456826783668105e-05,
+      "loss": 3.5805,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455988188917053e-05,
+      "loss": 3.5785,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4551512320463744e-05,
+      "loss": 3.5703,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4543126372953224e-05,
+      "loss": 3.5764,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4534740425442704e-05,
+      "loss": 3.5824,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4526354477932184e-05,
+      "loss": 3.5709,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45179849092254e-05,
+      "loss": 3.5933,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450959896171487e-05,
+      "loss": 3.5831,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450121301420435e-05,
+      "loss": 3.567,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449284344549757e-05,
+      "loss": 3.5958,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448445749798705e-05,
+      "loss": 3.5747,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447607155047652e-05,
+      "loss": 3.5699,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4467685602966e-05,
+      "loss": 3.5823,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445931603425922e-05,
+      "loss": 3.5789,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44509300867487e-05,
+      "loss": 3.5712,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444254413923818e-05,
+      "loss": 3.583,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443415819172766e-05,
+      "loss": 3.5681,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442578862302087e-05,
+      "loss": 3.5718,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4417402675510346e-05,
+      "loss": 3.5879,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4409016727999826e-05,
+      "loss": 3.5813,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4400630780489306e-05,
+      "loss": 3.5885,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439226121178252e-05,
+      "loss": 3.5811,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4383875264271995e-05,
+      "loss": 3.5967,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4375489316761475e-05,
+      "loss": 3.5714,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4367103369250955e-05,
+      "loss": 3.5916,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435873380054417e-05,
+      "loss": 3.5772,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435034785303365e-05,
+      "loss": 3.5722,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434196190552313e-05,
+      "loss": 3.5821,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433357595801261e-05,
+      "loss": 3.5931,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432520638930582e-05,
+      "loss": 3.5812,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43168204417953e-05,
+      "loss": 3.59,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430843449428478e-05,
+      "loss": 3.5662,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430004854677426e-05,
+      "loss": 3.5624,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429167897806747e-05,
+      "loss": 3.5715,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428329303055695e-05,
+      "loss": 3.5814,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427490708304643e-05,
+      "loss": 3.5783,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426652113553591e-05,
+      "loss": 3.578,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4258151566829125e-05,
+      "loss": 3.5713,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4249765619318605e-05,
+      "loss": 3.5796,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4241379671808085e-05,
+      "loss": 3.5738,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4232993724297565e-05,
+      "loss": 3.5743,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4224624155590774e-05,
+      "loss": 3.5862,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4216238208080254e-05,
+      "loss": 3.5896,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4207852260569734e-05,
+      "loss": 3.586,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4199466313059214e-05,
+      "loss": 3.5684,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419109674435242e-05,
+      "loss": 3.5766,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41827107968419e-05,
+      "loss": 3.5722,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417432484933138e-05,
+      "loss": 3.5838,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416593890182086e-05,
+      "loss": 3.5838,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415756933311408e-05,
+      "loss": 3.5774,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414918338560356e-05,
+      "loss": 3.5768,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414079743809304e-05,
+      "loss": 3.5793,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413241149058252e-05,
+      "loss": 3.5805,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.412404192187573e-05,
+      "loss": 3.5736,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411565597436521e-05,
+      "loss": 3.5739,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410727002685469e-05,
+      "loss": 3.5698,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409888407934417e-05,
+      "loss": 3.5722,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4090514510637376e-05,
+      "loss": 3.5714,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4082128563126856e-05,
+      "loss": 3.5704,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4073742615616336e-05,
+      "loss": 3.5746,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4065356668105816e-05,
+      "loss": 3.579,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405698709939903e-05,
+      "loss": 3.5663,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404860115188851e-05,
+      "loss": 3.5849,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404021520437799e-05,
+      "loss": 3.5909,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403182925686747e-05,
+      "loss": 3.5839,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402345968816068e-05,
+      "loss": 3.5809,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401507374065016e-05,
+      "loss": 3.5789,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400668779313964e-05,
+      "loss": 3.5699,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399830184562912e-05,
+      "loss": 3.5824,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398993227692233e-05,
+      "loss": 3.5605,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398154632941181e-05,
+      "loss": 3.5756,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397316038190129e-05,
+      "loss": 3.5764,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396477443439077e-05,
+      "loss": 3.5813,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3956404865683986e-05,
+      "loss": 3.5741,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3948018918173466e-05,
+      "loss": 3.5822,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3939632970662946e-05,
+      "loss": 3.5706,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3931263401956155e-05,
+      "loss": 3.5696,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3922877454445635e-05,
+      "loss": 3.5817,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3914491506935115e-05,
+      "loss": 3.5808,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3906105559424595e-05,
+      "loss": 3.5895,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3897735990717804e-05,
+      "loss": 3.5736,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3889350043207284e-05,
+      "loss": 3.5861,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3880964095696764e-05,
+      "loss": 3.5758,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3872578148186244e-05,
+      "loss": 3.5803,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386420857947945e-05,
+      "loss": 3.5825,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385582263196894e-05,
+      "loss": 3.5751,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384745306326215e-05,
+      "loss": 3.5731,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383906711575163e-05,
+      "loss": 3.5885,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383068116824111e-05,
+      "loss": 3.5675,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382229522073059e-05,
+      "loss": 3.5764,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381390927322007e-05,
+      "loss": 3.581,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380552332570955e-05,
+      "loss": 3.5791,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379713737819903e-05,
+      "loss": 3.5758,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378875143068851e-05,
+      "loss": 3.5766,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378038186198172e-05,
+      "loss": 3.5782,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37719959144712e-05,
+      "loss": 3.5839,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376360996696068e-05,
+      "loss": 3.5798,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8408312797546387,
+      "eval_runtime": 344.4653,
+      "eval_samples_per_second": 1107.778,
+      "eval_steps_per_second": 34.619,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375522401945016e-05,
+      "loss": 3.5683,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374685445074337e-05,
+      "loss": 3.5622,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373846850323285e-05,
+      "loss": 3.5728,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373008255572233e-05,
+      "loss": 3.5885,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372169660821181e-05,
+      "loss": 3.5801,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371332703950502e-05,
+      "loss": 3.5759,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.37049410919945e-05,
+      "loss": 3.5766,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369655514448398e-05,
+      "loss": 3.5723,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368816919697346e-05,
+      "loss": 3.5637,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367979962826667e-05,
+      "loss": 3.5669,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367141368075615e-05,
+      "loss": 3.5778,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366302773324563e-05,
+      "loss": 3.5819,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365464178573511e-05,
+      "loss": 3.5842,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364627221702833e-05,
+      "loss": 3.552,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363788626951781e-05,
+      "loss": 3.5697,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362950032200729e-05,
+      "loss": 3.5583,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362111437449677e-05,
+      "loss": 3.5708,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3612744805789976e-05,
+      "loss": 3.5702,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3604358858279456e-05,
+      "loss": 3.5721,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3595972910768936e-05,
+      "loss": 3.568,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3587586963258416e-05,
+      "loss": 3.5886,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3579217394551625e-05,
+      "loss": 3.5697,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3570831447041105e-05,
+      "loss": 3.5744,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3562445499530585e-05,
+      "loss": 3.568,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3554059552020065e-05,
+      "loss": 3.5734,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354568998331328e-05,
+      "loss": 3.5736,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353730403580276e-05,
+      "loss": 3.5655,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352891808829224e-05,
+      "loss": 3.5704,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352054851958545e-05,
+      "loss": 3.5626,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351216257207493e-05,
+      "loss": 3.5637,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350377662456441e-05,
+      "loss": 3.5729,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349539067705389e-05,
+      "loss": 3.5691,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34870211083471e-05,
+      "loss": 3.5736,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347863516083658e-05,
+      "loss": 3.5754,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347024921332606e-05,
+      "loss": 3.5713,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346186326581554e-05,
+      "loss": 3.5691,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3453493697108754e-05,
+      "loss": 3.5806,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3445107749598234e-05,
+      "loss": 3.5544,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3436721802087714e-05,
+      "loss": 3.5639,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3428335854577194e-05,
+      "loss": 3.5684,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34199662858704e-05,
+      "loss": 3.5573,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341158033835988e-05,
+      "loss": 3.5684,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340319439084936e-05,
+      "loss": 3.5699,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339480844333884e-05,
+      "loss": 3.5676,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338643887463205e-05,
+      "loss": 3.5707,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337805292712153e-05,
+      "loss": 3.5708,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336966697961101e-05,
+      "loss": 3.5767,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336128103210049e-05,
+      "loss": 3.5682,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335291146339371e-05,
+      "loss": 3.5498,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334452551588319e-05,
+      "loss": 3.5702,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333613956837267e-05,
+      "loss": 3.5667,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332775362086214e-05,
+      "loss": 3.5713,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331938405215536e-05,
+      "loss": 3.5637,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331099810464484e-05,
+      "loss": 3.5588,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330261215713432e-05,
+      "loss": 3.5551,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.329422620962379e-05,
+      "loss": 3.5598,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3285856640917006e-05,
+      "loss": 3.5692,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3277470693406486e-05,
+      "loss": 3.5519,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326908474589596e-05,
+      "loss": 3.5776,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3260698798385446e-05,
+      "loss": 3.5692,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3252329229678655e-05,
+      "loss": 3.5481,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324394328216814e-05,
+      "loss": 3.5741,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3235557334657615e-05,
+      "loss": 3.5576,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3227171387147095e-05,
+      "loss": 3.5581,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321880181844031e-05,
+      "loss": 3.5579,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321041587092979e-05,
+      "loss": 3.5653,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3202029923419264e-05,
+      "loss": 3.554,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3193643975908744e-05,
+      "loss": 3.568,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318527440720196e-05,
+      "loss": 3.5476,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317688845969143e-05,
+      "loss": 3.5565,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316850251218091e-05,
+      "loss": 3.5723,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316011656467039e-05,
+      "loss": 3.5617,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315174699596361e-05,
+      "loss": 3.5713,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314336104845309e-05,
+      "loss": 3.5638,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313497510094257e-05,
+      "loss": 3.5805,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312658915343205e-05,
+      "loss": 3.5549,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3118219584725264e-05,
+      "loss": 3.5747,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310983363721474e-05,
+      "loss": 3.5585,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310144768970422e-05,
+      "loss": 3.5583,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30930617421937e-05,
+      "loss": 3.5652,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3084692173486906e-05,
+      "loss": 3.576,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3076306225976386e-05,
+      "loss": 3.566,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3067920278465866e-05,
+      "loss": 3.5701,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3059534330955346e-05,
+      "loss": 3.5491,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305116476224856e-05,
+      "loss": 3.5486,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304277881473804e-05,
+      "loss": 3.5576,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303439286722752e-05,
+      "loss": 3.5613,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3026006919717e-05,
+      "loss": 3.5604,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301763735101021e-05,
+      "loss": 3.5603,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.300925140349969e-05,
+      "loss": 3.5604,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.300086545598917e-05,
+      "loss": 3.5616,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299247950847865e-05,
+      "loss": 3.5598,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298410993977186e-05,
+      "loss": 3.558,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297572399226134e-05,
+      "loss": 3.5673,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296733804475082e-05,
+      "loss": 3.5734,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29589520972403e-05,
+      "loss": 3.5685,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2950582528533516e-05,
+      "loss": 3.5536,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2942196581022996e-05,
+      "loss": 3.5578,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2933810633512476e-05,
+      "loss": 3.5564,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2925441064805685e-05,
+      "loss": 3.5675,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2917055117295165e-05,
+      "loss": 3.564,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2908669169784645e-05,
+      "loss": 3.5605,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2900283222274125e-05,
+      "loss": 3.5594,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2891913653567334e-05,
+      "loss": 3.5657,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2883527706056814e-05,
+      "loss": 3.5606,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2875141758546294e-05,
+      "loss": 3.5603,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2866755811035774e-05,
+      "loss": 3.5553,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285838624232899e-05,
+      "loss": 3.5545,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285000029481847e-05,
+      "loss": 3.561,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284161434730795e-05,
+      "loss": 3.5521,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283322839979743e-05,
+      "loss": 3.5561,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.282485883109064e-05,
+      "loss": 3.5574,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281647288358012e-05,
+      "loss": 3.5607,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28080869360696e-05,
+      "loss": 3.5559,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279970098855908e-05,
+      "loss": 3.5645,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279133141985229e-05,
+      "loss": 3.5739,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278294547234177e-05,
+      "loss": 3.5677,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277455952483125e-05,
+      "loss": 3.5665,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276617357732073e-05,
+      "loss": 3.559,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275780400861394e-05,
+      "loss": 3.5588,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274941806110342e-05,
+      "loss": 3.5613,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27410321135929e-05,
+      "loss": 3.5468,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273264616608238e-05,
+      "loss": 3.5603,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272427659737559e-05,
+      "loss": 3.5539,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271589064986507e-05,
+      "loss": 3.5661,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270750470235455e-05,
+      "loss": 3.5603,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269911875484403e-05,
+      "loss": 3.564,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269074918613724e-05,
+      "loss": 3.5526,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268236323862672e-05,
+      "loss": 3.5566,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26739772911162e-05,
+      "loss": 3.565,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266559134360568e-05,
+      "loss": 3.5657,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26572217748989e-05,
+      "loss": 3.573,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264883582738838e-05,
+      "loss": 3.5559,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264044987987786e-05,
+      "loss": 3.5675,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263206393236734e-05,
+      "loss": 3.5609,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2623694363660546e-05,
+      "loss": 3.5635,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2615308416150026e-05,
+      "loss": 3.5673,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2606922468639506e-05,
+      "loss": 3.5585,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2598536521128986e-05,
+      "loss": 3.5584,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2590166952422195e-05,
+      "loss": 3.5706,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2581781004911675e-05,
+      "loss": 3.5552,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2573395057401155e-05,
+      "loss": 3.5554,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2565009109890635e-05,
+      "loss": 3.5656,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255663954118385e-05,
+      "loss": 3.5638,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254825359367333e-05,
+      "loss": 3.5563,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253986764616281e-05,
+      "loss": 3.562,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253148169865229e-05,
+      "loss": 3.5569,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25231121299455e-05,
+      "loss": 3.5717,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251472618243498e-05,
+      "loss": 3.5601,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8400497436523438,
+      "eval_runtime": 343.4204,
+      "eval_samples_per_second": 1111.148,
+      "eval_steps_per_second": 34.724,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.250634023492446e-05,
+      "loss": 3.5438,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249795428741394e-05,
+      "loss": 3.548,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248956833990342e-05,
+      "loss": 3.557,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24811823923929e-05,
+      "loss": 3.5703,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247281282368611e-05,
+      "loss": 3.5631,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246442687617559e-05,
+      "loss": 3.56,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245604092866507e-05,
+      "loss": 3.5573,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.244765498115455e-05,
+      "loss": 3.5554,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2439285412447764e-05,
+      "loss": 3.5453,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2430899464937244e-05,
+      "loss": 3.5573,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2422513517426724e-05,
+      "loss": 3.5572,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2414127569916204e-05,
+      "loss": 3.571,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240575800120941e-05,
+      "loss": 3.5676,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239737205369889e-05,
+      "loss": 3.5358,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238898610618837e-05,
+      "loss": 3.5554,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238060015867785e-05,
+      "loss": 3.54,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237223058997106e-05,
+      "loss": 3.5543,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236384464246054e-05,
+      "loss": 3.5524,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235545869495002e-05,
+      "loss": 3.5588,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23470727474395e-05,
+      "loss": 3.5496,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.233870317873272e-05,
+      "loss": 3.5716,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23303172312222e-05,
+      "loss": 3.554,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232193128371168e-05,
+      "loss": 3.5591,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231354533620116e-05,
+      "loss": 3.5511,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230517576749437e-05,
+      "loss": 3.5631,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229678981998385e-05,
+      "loss": 3.5541,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228840387247333e-05,
+      "loss": 3.5504,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2280017924962807e-05,
+      "loss": 3.5542,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2271648356256016e-05,
+      "loss": 3.5511,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2263262408745496e-05,
+      "loss": 3.5482,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2254876461234976e-05,
+      "loss": 3.5552,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2246490513724456e-05,
+      "loss": 3.5518,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223812094501767e-05,
+      "loss": 3.5584,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222973499750715e-05,
+      "loss": 3.5567,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222134904999663e-05,
+      "loss": 3.5572,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221296310248611e-05,
+      "loss": 3.559,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220459353377932e-05,
+      "loss": 3.559,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21962075862688e-05,
+      "loss": 3.5416,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218782163875828e-05,
+      "loss": 3.549,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2179435691247753e-05,
+      "loss": 3.549,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217106612254097e-05,
+      "loss": 3.5414,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216268017503045e-05,
+      "loss": 3.5533,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215429422751993e-05,
+      "loss": 3.5576,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214590828000941e-05,
+      "loss": 3.5473,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2137538711302625e-05,
+      "loss": 3.5564,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2129152763792105e-05,
+      "loss": 3.555,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2120766816281585e-05,
+      "loss": 3.5603,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2112397247574794e-05,
+      "loss": 3.553,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2104011300064274e-05,
+      "loss": 3.5346,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2095625352553754e-05,
+      "loss": 3.5523,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208725578384696e-05,
+      "loss": 3.552,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207886983633644e-05,
+      "loss": 3.5556,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207048388882592e-05,
+      "loss": 3.5478,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20620979413154e-05,
+      "loss": 3.5427,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205371199380488e-05,
+      "loss": 3.5406,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204532604629436e-05,
+      "loss": 3.5399,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203694009878384e-05,
+      "loss": 3.5525,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202855415127332e-05,
+      "loss": 3.5354,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202018458256653e-05,
+      "loss": 3.5616,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201179863505601e-05,
+      "loss": 3.552,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200341268754549e-05,
+      "loss": 3.5326,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199502674003497e-05,
+      "loss": 3.559,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198665717132818e-05,
+      "loss": 3.5401,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197827122381766e-05,
+      "loss": 3.5436,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196988527630714e-05,
+      "loss": 3.5405,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196149932879662e-05,
+      "loss": 3.5489,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1953129760089837e-05,
+      "loss": 3.5403,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1944743812579317e-05,
+      "loss": 3.5514,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1936357865068797e-05,
+      "loss": 3.5293,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1927971917558276e-05,
+      "loss": 3.5434,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1919602348851486e-05,
+      "loss": 3.5531,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1911216401340966e-05,
+      "loss": 3.5472,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1902830453830445e-05,
+      "loss": 3.5536,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1894444506319925e-05,
+      "loss": 3.5483,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1886074937613135e-05,
+      "loss": 3.5645,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1877688990102614e-05,
+      "loss": 3.5429,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1869303042592094e-05,
+      "loss": 3.5594,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1860917095081574e-05,
+      "loss": 3.5409,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185254752637479e-05,
+      "loss": 3.5452,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1844177957668006e-05,
+      "loss": 3.5499,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183579201015748e-05,
+      "loss": 3.5564,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182740606264696e-05,
+      "loss": 3.5515,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181902011513644e-05,
+      "loss": 3.5541,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181063416762592e-05,
+      "loss": 3.5365,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18022482201154e-05,
+      "loss": 3.5291,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179386227260488e-05,
+      "loss": 3.5418,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178547632509436e-05,
+      "loss": 3.5473,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177710675638757e-05,
+      "loss": 3.5454,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176872080887705e-05,
+      "loss": 3.5416,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176033486136653e-05,
+      "loss": 3.5469,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175194891385601e-05,
+      "loss": 3.5425,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1743579345149224e-05,
+      "loss": 3.5442,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1735193397638704e-05,
+      "loss": 3.5441,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1726807450128184e-05,
+      "loss": 3.5532,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1718421502617664e-05,
+      "loss": 3.5573,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171005193391087e-05,
+      "loss": 3.5569,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170166598640035e-05,
+      "loss": 3.5354,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169328003888983e-05,
+      "loss": 3.5432,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168489409137931e-05,
+      "loss": 3.5424,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167652452267252e-05,
+      "loss": 3.5506,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1668138575162e-05,
+      "loss": 3.55,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165975262765148e-05,
+      "loss": 3.5445,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165136668014096e-05,
+      "loss": 3.5454,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164299711143418e-05,
+      "loss": 3.5473,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163462754272739e-05,
+      "loss": 3.5483,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1626241595216867e-05,
+      "loss": 3.5407,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1617855647706347e-05,
+      "loss": 3.546,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1609469700195827e-05,
+      "loss": 3.5353,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1601083752685306e-05,
+      "loss": 3.5445,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1592697805174786e-05,
+      "loss": 3.5353,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1584311857664266e-05,
+      "loss": 3.5408,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1575925910153746e-05,
+      "loss": 3.5404,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1567556341446955e-05,
+      "loss": 3.547,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1559170393936435e-05,
+      "loss": 3.5409,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1550784446425915e-05,
+      "loss": 3.5472,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1542398498915395e-05,
+      "loss": 3.5562,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153402893020861e-05,
+      "loss": 3.5547,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152564298269809e-05,
+      "loss": 3.552,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151725703518757e-05,
+      "loss": 3.5425,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150887108767705e-05,
+      "loss": 3.5425,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150050151897026e-05,
+      "loss": 3.5442,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149211557145974e-05,
+      "loss": 3.5298,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.148372962394922e-05,
+      "loss": 3.5472,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14753436764387e-05,
+      "loss": 3.5395,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146697410773191e-05,
+      "loss": 3.5531,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145858816022139e-05,
+      "loss": 3.542,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145020221271087e-05,
+      "loss": 3.5476,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144181626520035e-05,
+      "loss": 3.5398,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1433446696493565e-05,
+      "loss": 3.5411,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1425060748983045e-05,
+      "loss": 3.5502,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1416674801472525e-05,
+      "loss": 3.5471,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1408288853962005e-05,
+      "loss": 3.5604,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1399919285255214e-05,
+      "loss": 3.5418,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1391533337744694e-05,
+      "loss": 3.55,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1383147390234174e-05,
+      "loss": 3.5478,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1374761442723654e-05,
+      "loss": 3.5484,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136639187401686e-05,
+      "loss": 3.5477,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135800592650634e-05,
+      "loss": 3.5421,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134961997899582e-05,
+      "loss": 3.5497,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13412340314853e-05,
+      "loss": 3.5516,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133286446277852e-05,
+      "loss": 3.5414,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1324478515268e-05,
+      "loss": 3.5416,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131609256775748e-05,
+      "loss": 3.5476,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130770662024696e-05,
+      "loss": 3.5469,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129933705154017e-05,
+      "loss": 3.5407,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129095110402965e-05,
+      "loss": 3.5475,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128256515651913e-05,
+      "loss": 3.5422,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127417920900861e-05,
+      "loss": 3.5607,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1265809640301816e-05,
+      "loss": 3.5423,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8404266834259033,
+      "eval_runtime": 303.6737,
+      "eval_samples_per_second": 1256.582,
+      "eval_steps_per_second": 39.269,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1257423692791296e-05,
+      "loss": 3.5385,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1249037745280776e-05,
+      "loss": 3.5348,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1240651797770256e-05,
+      "loss": 3.5368,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123228222906347e-05,
+      "loss": 3.5533,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.122389628155295e-05,
+      "loss": 3.5485,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121551033404243e-05,
+      "loss": 3.5458,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120714076533564e-05,
+      "loss": 3.5399,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119875481782512e-05,
+      "loss": 3.5402,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11903688703146e-05,
+      "loss": 3.5343,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118198292280408e-05,
+      "loss": 3.5417,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117361335409729e-05,
+      "loss": 3.5399,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116522740658677e-05,
+      "loss": 3.5564,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115684145907625e-05,
+      "loss": 3.5514,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114845551156573e-05,
+      "loss": 3.5234,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1140085942858946e-05,
+      "loss": 3.5389,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1131699995348426e-05,
+      "loss": 3.527,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1123314047837906e-05,
+      "loss": 3.5372,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.111492810032738e-05,
+      "loss": 3.5371,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1106558531620595e-05,
+      "loss": 3.5432,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1098172584110075e-05,
+      "loss": 3.5353,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1089786636599555e-05,
+      "loss": 3.5559,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108140068908903e-05,
+      "loss": 3.5447,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1073031120382244e-05,
+      "loss": 3.5378,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1064645172871724e-05,
+      "loss": 3.5351,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1056259225361204e-05,
+      "loss": 3.5513,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1047873277850684e-05,
+      "loss": 3.5371,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.10395037091439e-05,
+      "loss": 3.5342,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103111776163338e-05,
+      "loss": 3.5424,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102273181412285e-05,
+      "loss": 3.5318,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.101434586661233e-05,
+      "loss": 3.533,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.100595991910181e-05,
+      "loss": 3.5394,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099757397159129e-05,
+      "loss": 3.5371,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098918802408077e-05,
+      "loss": 3.5422,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098080207657025e-05,
+      "loss": 3.5419,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097243250786346e-05,
+      "loss": 3.5452,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096404656035294e-05,
+      "loss": 3.5408,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095566061284242e-05,
+      "loss": 3.5451,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09472746653319e-05,
+      "loss": 3.5286,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093890509662512e-05,
+      "loss": 3.536,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09305191491146e-05,
+      "loss": 3.5328,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092213320160408e-05,
+      "loss": 3.5296,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091374725409356e-05,
+      "loss": 3.5357,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0905377685386766e-05,
+      "loss": 3.5431,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0896991737876246e-05,
+      "loss": 3.5306,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0888605790365726e-05,
+      "loss": 3.5409,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0880219842855206e-05,
+      "loss": 3.5426,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0871850274148415e-05,
+      "loss": 3.5436,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0863464326637895e-05,
+      "loss": 3.5383,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0855078379127375e-05,
+      "loss": 3.5186,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0846692431616855e-05,
+      "loss": 3.5365,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083832286291007e-05,
+      "loss": 3.5421,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082993691539955e-05,
+      "loss": 3.54,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082155096788903e-05,
+      "loss": 3.5291,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081318139918224e-05,
+      "loss": 3.5295,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.080479545167172e-05,
+      "loss": 3.5252,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07964095041612e-05,
+      "loss": 3.5217,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.078802355665068e-05,
+      "loss": 3.5408,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077965398794389e-05,
+      "loss": 3.5229,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077126804043337e-05,
+      "loss": 3.5432,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076288209292285e-05,
+      "loss": 3.5339,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075449614541233e-05,
+      "loss": 3.5251,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0746126576705545e-05,
+      "loss": 3.5421,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0737740629195025e-05,
+      "loss": 3.5223,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0729354681684505e-05,
+      "loss": 3.528,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0720968734173985e-05,
+      "loss": 3.5247,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0712599165467194e-05,
+      "loss": 3.5375,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0704213217956674e-05,
+      "loss": 3.5219,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0695827270446154e-05,
+      "loss": 3.5357,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0687441322935634e-05,
+      "loss": 3.5146,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067907175422884e-05,
+      "loss": 3.5277,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067068580671832e-05,
+      "loss": 3.5406,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06622998592078e-05,
+      "loss": 3.5359,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065393029050102e-05,
+      "loss": 3.5359,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06455443429905e-05,
+      "loss": 3.5382,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063715839547998e-05,
+      "loss": 3.5475,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062877244796946e-05,
+      "loss": 3.5288,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062038650045894e-05,
+      "loss": 3.5452,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061200055294842e-05,
+      "loss": 3.5233,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06036146054379e-05,
+      "loss": 3.5289,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059524503673111e-05,
+      "loss": 3.5344,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058685908922059e-05,
+      "loss": 3.5413,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057847314171007e-05,
+      "loss": 3.5336,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057008719419955e-05,
+      "loss": 3.5437,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056170124668903e-05,
+      "loss": 3.5253,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055331529917851e-05,
+      "loss": 3.5158,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.054492935166798e-05,
+      "loss": 3.5258,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.053654340415747e-05,
+      "loss": 3.5317,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052817383545068e-05,
+      "loss": 3.5275,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051978788794016e-05,
+      "loss": 3.5276,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0511401940429636e-05,
+      "loss": 3.5358,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0503015992919116e-05,
+      "loss": 3.528,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0494646424212332e-05,
+      "loss": 3.5289,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0486260476701805e-05,
+      "loss": 3.528,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0477874529191285e-05,
+      "loss": 3.5404,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0469488581680765e-05,
+      "loss": 3.541,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046111901297398e-05,
+      "loss": 3.5399,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0452733065463457e-05,
+      "loss": 3.5241,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0444347117952937e-05,
+      "loss": 3.5313,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0435961170442417e-05,
+      "loss": 3.5251,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0427591601735633e-05,
+      "loss": 3.5365,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041920565422511e-05,
+      "loss": 3.5354,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041081970671459e-05,
+      "loss": 3.5283,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040243375920407e-05,
+      "loss": 3.5303,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.039406419049728e-05,
+      "loss": 3.5351,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.038567824298676e-05,
+      "loss": 3.5331,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.037729229547624e-05,
+      "loss": 3.5263,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.036890634796572e-05,
+      "loss": 3.5305,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.036053677925893e-05,
+      "loss": 3.5216,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035215083174841e-05,
+      "loss": 3.5294,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.034376488423789e-05,
+      "loss": 3.5212,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.033537893672737e-05,
+      "loss": 3.5258,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.032700936802058e-05,
+      "loss": 3.5253,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0318623420510063e-05,
+      "loss": 3.5324,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0310237472999543e-05,
+      "loss": 3.5268,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0301851525489023e-05,
+      "loss": 3.5281,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0293481956782232e-05,
+      "loss": 3.5445,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0285096009271712e-05,
+      "loss": 3.5397,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0276710061761192e-05,
+      "loss": 3.5352,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0268324114250672e-05,
+      "loss": 3.5282,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0259954545543885e-05,
+      "loss": 3.5293,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0251568598033365e-05,
+      "loss": 3.5277,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0243182650522845e-05,
+      "loss": 3.5178,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0234813081816054e-05,
+      "loss": 3.5287,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0226427134305534e-05,
+      "loss": 3.5274,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0218041186795017e-05,
+      "loss": 3.5346,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0209655239284497e-05,
+      "loss": 3.5284,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0201285670577706e-05,
+      "loss": 3.5353,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0192899723067186e-05,
+      "loss": 3.5276,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0184513775556666e-05,
+      "loss": 3.526,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0176127828046146e-05,
+      "loss": 3.531,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.016775825933936e-05,
+      "loss": 3.5364,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.015937231182884e-05,
+      "loss": 3.5417,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0150986364318318e-05,
+      "loss": 3.5315,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0142600416807798e-05,
+      "loss": 3.5356,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0134230848101007e-05,
+      "loss": 3.5322,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0125844900590487e-05,
+      "loss": 3.535,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0117458953079967e-05,
+      "loss": 3.5327,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010907300556945e-05,
+      "loss": 3.529,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010070343686266e-05,
+      "loss": 3.5345,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009231748935214e-05,
+      "loss": 3.5359,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.008393154184162e-05,
+      "loss": 3.5279,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.00755455943311e-05,
+      "loss": 3.5277,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0067176025624312e-05,
+      "loss": 3.5323,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0058790078113792e-05,
+      "loss": 3.5355,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0050404130603272e-05,
+      "loss": 3.5237,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0042018183092752e-05,
+      "loss": 3.5378,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.003364861438596e-05,
+      "loss": 3.5257,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.002526266687544e-05,
+      "loss": 3.5419,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.001687671936492e-05,
+      "loss": 3.5298,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0008490771854404e-05,
+      "loss": 3.5269,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8418712615966797,
+      "eval_runtime": 304.4156,
+      "eval_samples_per_second": 1253.52,
+      "eval_steps_per_second": 39.173,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.0000104824343884e-05,
+      "loss": 3.5187,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9991718876833364e-05,
+      "loss": 3.5218,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9983332929322844e-05,
+      "loss": 3.5415,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9974963360616053e-05,
+      "loss": 3.5358,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9966577413105533e-05,
+      "loss": 3.5334,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9958191465595013e-05,
+      "loss": 3.5272,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9949805518084496e-05,
+      "loss": 3.5252,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9941435949377706e-05,
+      "loss": 3.5137,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9933050001867186e-05,
+      "loss": 3.5329,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9924664054356665e-05,
+      "loss": 3.5222,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9916278106846145e-05,
+      "loss": 3.5453,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9907908538139358e-05,
+      "loss": 3.5376,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9899522590628838e-05,
+      "loss": 3.5107,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9891136643118318e-05,
+      "loss": 3.5226,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9882750695607798e-05,
+      "loss": 3.5185,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9874381126901007e-05,
+      "loss": 3.5184,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9865995179390487e-05,
+      "loss": 3.5271,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9857609231879967e-05,
+      "loss": 3.5301,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984922328436945e-05,
+      "loss": 3.5184,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984085371566266e-05,
+      "loss": 3.5446,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983246776815214e-05,
+      "loss": 3.5279,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.982408182064162e-05,
+      "loss": 3.5285,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98156958731311e-05,
+      "loss": 3.5219,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.980732630442431e-05,
+      "loss": 3.5345,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.979894035691379e-05,
+      "loss": 3.5235,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9790570788207e-05,
+      "loss": 3.5205,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978218484069648e-05,
+      "loss": 3.5274,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.977379889318596e-05,
+      "loss": 3.5198,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.976541294567544e-05,
+      "loss": 3.5181,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.975702699816492e-05,
+      "loss": 3.5217,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9748641050654404e-05,
+      "loss": 3.527,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9740255103143884e-05,
+      "loss": 3.5274,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9731869155633357e-05,
+      "loss": 3.5271,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9723499586926573e-05,
+      "loss": 3.5308,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9715113639416053e-05,
+      "loss": 3.5278,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9706727691905533e-05,
+      "loss": 3.53,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.969834174439501e-05,
+      "loss": 3.5162,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9689972175688225e-05,
+      "loss": 3.5221,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9681586228177705e-05,
+      "loss": 3.5161,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9673200280667185e-05,
+      "loss": 3.5205,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9664814333156658e-05,
+      "loss": 3.52,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9656444764449874e-05,
+      "loss": 3.5295,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9648058816939357e-05,
+      "loss": 3.5189,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963967286942883e-05,
+      "loss": 3.5273,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963128692191831e-05,
+      "loss": 3.5253,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9622917353211526e-05,
+      "loss": 3.5332,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9614531405701006e-05,
+      "loss": 3.5226,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9606145458190483e-05,
+      "loss": 3.5089,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9597759510679963e-05,
+      "loss": 3.5163,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.958938994197318e-05,
+      "loss": 3.5341,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9581003994462652e-05,
+      "loss": 3.5236,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9572618046952132e-05,
+      "loss": 3.5135,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9564232099441612e-05,
+      "loss": 3.5114,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9555862530734828e-05,
+      "loss": 3.5166,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9547476583224304e-05,
+      "loss": 3.5079,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9539090635713784e-05,
+      "loss": 3.5301,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9530721067007e-05,
+      "loss": 3.5097,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.952233511949648e-05,
+      "loss": 3.5239,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9513949171985957e-05,
+      "loss": 3.521,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9505563224475437e-05,
+      "loss": 3.5086,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9497193655768653e-05,
+      "loss": 3.5285,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9488807708258126e-05,
+      "loss": 3.5109,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9480421760747606e-05,
+      "loss": 3.5187,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9472035813237086e-05,
+      "loss": 3.5073,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.94636662445303e-05,
+      "loss": 3.5257,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9455280297019778e-05,
+      "loss": 3.5123,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9446894349509258e-05,
+      "loss": 3.5137,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9438508401998738e-05,
+      "loss": 3.5067,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9430138833291954e-05,
+      "loss": 3.5147,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9421752885781427e-05,
+      "loss": 3.5231,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941336693827091e-05,
+      "loss": 3.527,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.940498099076039e-05,
+      "loss": 3.521,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.93966114220536e-05,
+      "loss": 3.5217,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.938822547454308e-05,
+      "loss": 3.5308,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.937983952703256e-05,
+      "loss": 3.5149,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.937145357952204e-05,
+      "loss": 3.5315,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9363084010815252e-05,
+      "loss": 3.5133,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9354698063304732e-05,
+      "loss": 3.514,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9346328494597948e-05,
+      "loss": 3.521,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9337942547087428e-05,
+      "loss": 3.5249,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.93295565995769e-05,
+      "loss": 3.5208,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.932117065206638e-05,
+      "loss": 3.5274,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931278470455586e-05,
+      "loss": 3.5159,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9304398757045344e-05,
+      "loss": 3.498,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9296012809534824e-05,
+      "loss": 3.511,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9287626862024304e-05,
+      "loss": 3.5169,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9279257293317513e-05,
+      "loss": 3.5138,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9270871345806993e-05,
+      "loss": 3.5154,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9262485398296473e-05,
+      "loss": 3.5228,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9254099450785953e-05,
+      "loss": 3.512,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9245729882079165e-05,
+      "loss": 3.5162,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9237343934568645e-05,
+      "loss": 3.5146,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9228957987058125e-05,
+      "loss": 3.5241,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9220572039547605e-05,
+      "loss": 3.5274,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9212202470840814e-05,
+      "loss": 3.5301,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9203816523330298e-05,
+      "loss": 3.5044,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9195430575819778e-05,
+      "loss": 3.517,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9187044628309258e-05,
+      "loss": 3.5147,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9178675059602467e-05,
+      "loss": 3.5179,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9170289112091947e-05,
+      "loss": 3.5248,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9161903164581427e-05,
+      "loss": 3.515,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9153517217070907e-05,
+      "loss": 3.514,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.914514764836412e-05,
+      "loss": 3.5216,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91367617008536e-05,
+      "loss": 3.5198,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9128392132146808e-05,
+      "loss": 3.5123,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9120006184636288e-05,
+      "loss": 3.517,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9111620237125768e-05,
+      "loss": 3.5115,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910323428961525e-05,
+      "loss": 3.5126,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.909484834210473e-05,
+      "loss": 3.5093,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.908646239459421e-05,
+      "loss": 3.5127,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.907807644708369e-05,
+      "loss": 3.5112,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.90697068783769e-05,
+      "loss": 3.5199,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.906132093086638e-05,
+      "loss": 3.5141,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905293498335586e-05,
+      "loss": 3.5145,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9044549035845344e-05,
+      "loss": 3.5316,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9036179467138553e-05,
+      "loss": 3.529,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9027793519628033e-05,
+      "loss": 3.5188,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9019407572117513e-05,
+      "loss": 3.5185,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9011021624606993e-05,
+      "loss": 3.5142,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9002652055900205e-05,
+      "loss": 3.5131,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8994266108389685e-05,
+      "loss": 3.5076,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8985880160879165e-05,
+      "loss": 3.5137,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8977494213368645e-05,
+      "loss": 3.5155,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8969124644661854e-05,
+      "loss": 3.52,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8960738697151334e-05,
+      "loss": 3.515,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8952352749640814e-05,
+      "loss": 3.5212,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8943966802130297e-05,
+      "loss": 3.5143,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8935597233423506e-05,
+      "loss": 3.5136,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8927211285912986e-05,
+      "loss": 3.512,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8918825338402466e-05,
+      "loss": 3.5278,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8910439390891946e-05,
+      "loss": 3.5278,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.890206982218516e-05,
+      "loss": 3.5144,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.889368387467464e-05,
+      "loss": 3.518,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.888529792716412e-05,
+      "loss": 3.5233,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.88769119796536e-05,
+      "loss": 3.5214,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8868542410946808e-05,
+      "loss": 3.5219,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8860156463436288e-05,
+      "loss": 3.514,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8851770515925768e-05,
+      "loss": 3.5229,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.884338456841525e-05,
+      "loss": 3.5176,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.883501499970846e-05,
+      "loss": 3.5158,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882662905219794e-05,
+      "loss": 3.5181,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881824310468742e-05,
+      "loss": 3.5165,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.88098571571769e-05,
+      "loss": 3.5219,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8801487588470112e-05,
+      "loss": 3.5106,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8793101640959592e-05,
+      "loss": 3.5236,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8784715693449072e-05,
+      "loss": 3.5163,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8776329745938552e-05,
+      "loss": 3.5244,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.876796017723176e-05,
+      "loss": 3.5211,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.875957422972124e-05,
+      "loss": 3.5095,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.842674970626831,
+      "eval_runtime": 301.8291,
+      "eval_samples_per_second": 1264.262,
+      "eval_steps_per_second": 39.509,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.875118828221072e-05,
+      "loss": 3.5043,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8742802334700205e-05,
+      "loss": 3.5087,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8734416387189678e-05,
+      "loss": 3.5308,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8726030439679158e-05,
+      "loss": 3.515,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8717660870972374e-05,
+      "loss": 3.5244,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8709274923461854e-05,
+      "loss": 3.514,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.870088897595133e-05,
+      "loss": 3.5105,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8692519407244546e-05,
+      "loss": 3.5033,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8684133459734026e-05,
+      "loss": 3.5171,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8675747512223506e-05,
+      "loss": 3.5087,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.866736156471298e-05,
+      "loss": 3.5311,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8658991996006195e-05,
+      "loss": 3.5219,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8650606048495675e-05,
+      "loss": 3.4994,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.864222010098515e-05,
+      "loss": 3.5072,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.863383415347463e-05,
+      "loss": 3.5088,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8625464584767847e-05,
+      "loss": 3.4981,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8617078637257327e-05,
+      "loss": 3.517,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8608692689746804e-05,
+      "loss": 3.5156,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8600306742236284e-05,
+      "loss": 3.5045,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.85919371735295e-05,
+      "loss": 3.5306,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858355122601898e-05,
+      "loss": 3.5139,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8575165278508453e-05,
+      "loss": 3.5164,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8566779330997933e-05,
+      "loss": 3.5076,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8558393383487413e-05,
+      "loss": 3.5232,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8550007435976893e-05,
+      "loss": 3.5033,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8541637867270105e-05,
+      "loss": 3.5125,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8533251919759585e-05,
+      "loss": 3.5104,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8524865972249065e-05,
+      "loss": 3.5074,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8516480024738545e-05,
+      "loss": 3.5035,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8508094077228025e-05,
+      "loss": 3.5124,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8499708129717505e-05,
+      "loss": 3.5143,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8491322182206985e-05,
+      "loss": 3.5125,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8482936234696468e-05,
+      "loss": 3.5124,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8474566665989677e-05,
+      "loss": 3.5201,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8466180718479157e-05,
+      "loss": 3.5116,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8457794770968637e-05,
+      "loss": 3.5158,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8449408823458117e-05,
+      "loss": 3.5049,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.844103925475133e-05,
+      "loss": 3.5093,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843265330724081e-05,
+      "loss": 3.5007,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.842426735973029e-05,
+      "loss": 3.5079,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.841588141221977e-05,
+      "loss": 3.5065,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.840751184351298e-05,
+      "loss": 3.5161,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.839914227480619e-05,
+      "loss": 3.5002,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.839075632729567e-05,
+      "loss": 3.5181,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.838237037978515e-05,
+      "loss": 3.5116,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.837398443227463e-05,
+      "loss": 3.518,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.836559848476411e-05,
+      "loss": 3.5083,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.835721253725359e-05,
+      "loss": 3.4966,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.834882658974307e-05,
+      "loss": 3.5042,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.834044064223255e-05,
+      "loss": 3.5171,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8332071073525763e-05,
+      "loss": 3.5103,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8323685126015243e-05,
+      "loss": 3.501,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8315299178504723e-05,
+      "loss": 3.498,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8306913230994203e-05,
+      "loss": 3.5068,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8298543662287412e-05,
+      "loss": 3.4898,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8290157714776892e-05,
+      "loss": 3.5197,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8281771767266376e-05,
+      "loss": 3.4949,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8273385819755855e-05,
+      "loss": 3.5118,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8265016251049065e-05,
+      "loss": 3.5059,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8256630303538545e-05,
+      "loss": 3.4988,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8248244356028024e-05,
+      "loss": 3.5097,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8239858408517504e-05,
+      "loss": 3.5025,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8231488839810717e-05,
+      "loss": 3.5036,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8223102892300197e-05,
+      "loss": 3.4925,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8214716944789677e-05,
+      "loss": 3.5161,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8206330997279157e-05,
+      "loss": 3.4971,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8197961428572366e-05,
+      "loss": 3.5051,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8189575481061846e-05,
+      "loss": 3.4889,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.818118953355133e-05,
+      "loss": 3.5035,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.817280358604081e-05,
+      "loss": 3.5087,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8164434017334018e-05,
+      "loss": 3.5106,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8156064448627227e-05,
+      "loss": 3.5085,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8147678501116707e-05,
+      "loss": 3.5121,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.813929255360619e-05,
+      "loss": 3.5141,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.813090660609567e-05,
+      "loss": 3.5058,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812252065858515e-05,
+      "loss": 3.5141,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.811413471107463e-05,
+      "loss": 3.5031,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.810574876356411e-05,
+      "loss": 3.504,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.809737919485732e-05,
+      "loss": 3.5047,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.80889932473468e-05,
+      "loss": 3.5107,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8080607299836283e-05,
+      "loss": 3.5103,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8072221352325756e-05,
+      "loss": 3.5146,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8063835404815236e-05,
+      "loss": 3.5066,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8055449457304716e-05,
+      "loss": 3.487,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8047063509794196e-05,
+      "loss": 3.4963,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8038677562283676e-05,
+      "loss": 3.5018,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.803030799357689e-05,
+      "loss": 3.5027,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8021922046066368e-05,
+      "loss": 3.4977,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8013536098555848e-05,
+      "loss": 3.5113,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8005150151045328e-05,
+      "loss": 3.497,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7996780582338537e-05,
+      "loss": 3.5004,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.798839463482802e-05,
+      "loss": 3.5029,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.79800086873175e-05,
+      "loss": 3.5121,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.797162273980698e-05,
+      "loss": 3.5116,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.796325317110019e-05,
+      "loss": 3.5176,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.795486722358967e-05,
+      "loss": 3.4916,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.794648127607915e-05,
+      "loss": 3.504,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.793809532856863e-05,
+      "loss": 3.5043,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7929725759861842e-05,
+      "loss": 3.5024,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7921339812351322e-05,
+      "loss": 3.514,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7912953864840802e-05,
+      "loss": 3.5037,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7904567917330282e-05,
+      "loss": 3.5014,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.789619834862349e-05,
+      "loss": 3.5108,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7887812401112974e-05,
+      "loss": 3.5082,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7879426453602454e-05,
+      "loss": 3.4957,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7871040506091934e-05,
+      "loss": 3.5069,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7862670937385143e-05,
+      "loss": 3.4997,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7854284989874623e-05,
+      "loss": 3.499,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7845899042364103e-05,
+      "loss": 3.4946,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7837513094853583e-05,
+      "loss": 3.4974,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7829143526146796e-05,
+      "loss": 3.4966,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7820757578636276e-05,
+      "loss": 3.5088,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7812371631125756e-05,
+      "loss": 3.4982,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7803985683615236e-05,
+      "loss": 3.5016,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7795616114908445e-05,
+      "loss": 3.5198,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7787230167397925e-05,
+      "loss": 3.5119,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7778844219887408e-05,
+      "loss": 3.5069,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7770458272376888e-05,
+      "loss": 3.5061,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7762088703670097e-05,
+      "loss": 3.5001,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7753702756159577e-05,
+      "loss": 3.5001,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7745316808649057e-05,
+      "loss": 3.4962,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.773694723994227e-05,
+      "loss": 3.4979,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.772856129243175e-05,
+      "loss": 3.5035,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.772017534492123e-05,
+      "loss": 3.5081,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.771178939741071e-05,
+      "loss": 3.5058,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.770341982870392e-05,
+      "loss": 3.5035,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76950338811934e-05,
+      "loss": 3.5027,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7686647933682878e-05,
+      "loss": 3.4962,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.767826198617236e-05,
+      "loss": 3.4983,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.766989241746557e-05,
+      "loss": 3.5152,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.766150646995505e-05,
+      "loss": 3.5141,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.765312052244453e-05,
+      "loss": 3.5093,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.764473457493401e-05,
+      "loss": 3.5026,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7636365006227223e-05,
+      "loss": 3.5051,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7627979058716703e-05,
+      "loss": 3.5058,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7619593111206183e-05,
+      "loss": 3.5127,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7611207163695663e-05,
+      "loss": 3.504,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7602837594988872e-05,
+      "loss": 3.5078,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7594451647478352e-05,
+      "loss": 3.5034,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7586065699967832e-05,
+      "loss": 3.5049,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7577679752457315e-05,
+      "loss": 3.501,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7569310183750524e-05,
+      "loss": 3.5041,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7560940615043733e-05,
+      "loss": 3.5091,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7552554667533213e-05,
+      "loss": 3.4947,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7544168720022693e-05,
+      "loss": 3.5137,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7535782772512177e-05,
+      "loss": 3.5066,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7527396825001657e-05,
+      "loss": 3.506,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7519010877491137e-05,
+      "loss": 3.5051,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7510624929980617e-05,
+      "loss": 3.4963,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.84421443939209,
+      "eval_runtime": 304.8143,
+      "eval_samples_per_second": 1251.88,
+      "eval_steps_per_second": 39.122,
+      "step": 1373760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7502238982470097e-05,
+      "loss": 3.4977,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7493869413763306e-05,
+      "loss": 3.4952,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7485483466252786e-05,
+      "loss": 3.5149,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.747709751874227e-05,
+      "loss": 3.5025,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.746871157123175e-05,
+      "loss": 3.5145,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7460342002524958e-05,
+      "loss": 3.4977,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7451956055014438e-05,
+      "loss": 3.5009,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7443570107503918e-05,
+      "loss": 3.491,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7435184159993398e-05,
+      "loss": 3.5037,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.742681459128661e-05,
+      "loss": 3.4964,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.741842864377609e-05,
+      "loss": 3.5149,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.741004269626557e-05,
+      "loss": 3.5082,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.740165674875505e-05,
+      "loss": 3.4887,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739328718004826e-05,
+      "loss": 3.4931,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.738490123253774e-05,
+      "loss": 3.5004,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7376515285027223e-05,
+      "loss": 3.4847,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7368129337516703e-05,
+      "loss": 3.5,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.735975976880991e-05,
+      "loss": 3.504,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.735137382129939e-05,
+      "loss": 3.4925,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.73430042525926e-05,
+      "loss": 3.5197,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7334618305082084e-05,
+      "loss": 3.4995,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7326232357571564e-05,
+      "loss": 3.5042,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7317846410061044e-05,
+      "loss": 3.4964,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7309460462550524e-05,
+      "loss": 3.5097,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7301074515040004e-05,
+      "loss": 3.4888,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7292704946333213e-05,
+      "loss": 3.5005,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7284318998822693e-05,
+      "loss": 3.4925,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7275933051312176e-05,
+      "loss": 3.4978,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7267547103801656e-05,
+      "loss": 3.4906,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7259161156291136e-05,
+      "loss": 3.4966,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.725077520878061e-05,
+      "loss": 3.5028,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.724238926127009e-05,
+      "loss": 3.4993,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.723400331375957e-05,
+      "loss": 3.4966,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7225633745052782e-05,
+      "loss": 3.511,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7217247797542262e-05,
+      "loss": 3.4956,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.720886185003174e-05,
+      "loss": 3.5028,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.720047590252122e-05,
+      "loss": 3.4915,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.719210633381443e-05,
+      "loss": 3.4969,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7183720386303914e-05,
+      "loss": 3.4848,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7175334438793394e-05,
+      "loss": 3.498,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7166948491282874e-05,
+      "loss": 3.4914,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7158578922576083e-05,
+      "loss": 3.5064,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7150192975065563e-05,
+      "loss": 3.4898,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.714182340635878e-05,
+      "loss": 3.5,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7133437458848255e-05,
+      "loss": 3.5045,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7125051511337735e-05,
+      "loss": 3.5016,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7116665563827215e-05,
+      "loss": 3.4995,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7108279616316695e-05,
+      "loss": 3.4849,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7099893668806175e-05,
+      "loss": 3.4905,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7091507721295655e-05,
+      "loss": 3.5039,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7083138152588868e-05,
+      "loss": 3.4976,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7074752205078348e-05,
+      "loss": 3.4919,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7066366257567828e-05,
+      "loss": 3.4845,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7057980310057308e-05,
+      "loss": 3.4966,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7049610741350517e-05,
+      "loss": 3.4758,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7041224793839997e-05,
+      "loss": 3.5079,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7032838846329477e-05,
+      "loss": 3.4825,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7024452898818957e-05,
+      "loss": 3.498,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.701608333011217e-05,
+      "loss": 3.495,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.700769738260165e-05,
+      "loss": 3.4856,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.699931143509113e-05,
+      "loss": 3.5008,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.699092548758061e-05,
+      "loss": 3.488,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698255591887382e-05,
+      "loss": 3.4902,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.69741699713633e-05,
+      "loss": 3.4765,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.696578402385278e-05,
+      "loss": 3.5022,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.695739807634226e-05,
+      "loss": 3.4863,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.694902850763547e-05,
+      "loss": 3.4898,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.694064256012495e-05,
+      "loss": 3.4783,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.693225661261443e-05,
+      "loss": 3.4891,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.692387066510391e-05,
+      "loss": 3.4952,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6915501096397123e-05,
+      "loss": 3.5002,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6907131527690332e-05,
+      "loss": 3.4899,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6898745580179812e-05,
+      "loss": 3.5026,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6890359632669292e-05,
+      "loss": 3.5003,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.688197368515877e-05,
+      "loss": 3.4954,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6873587737648255e-05,
+      "loss": 3.4963,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6865201790137735e-05,
+      "loss": 3.4966,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6856815842627215e-05,
+      "loss": 3.4897,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6848446273920424e-05,
+      "loss": 3.4923,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6840060326409904e-05,
+      "loss": 3.4971,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6831674378899384e-05,
+      "loss": 3.5003,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6823288431388864e-05,
+      "loss": 3.5039,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6814902483878347e-05,
+      "loss": 3.4907,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6806516536367827e-05,
+      "loss": 3.4717,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6798130588857307e-05,
+      "loss": 3.4843,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6789744641346787e-05,
+      "loss": 3.491,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6781375072639996e-05,
+      "loss": 3.4922,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6772989125129476e-05,
+      "loss": 3.4853,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.676461955642269e-05,
+      "loss": 3.4964,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.675623360891217e-05,
+      "loss": 3.4888,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.674784766140165e-05,
+      "loss": 3.4798,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.673946171389113e-05,
+      "loss": 3.4921,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.673107576638061e-05,
+      "loss": 3.5006,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.672268981887009e-05,
+      "loss": 3.4963,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6714303871359565e-05,
+      "loss": 3.5028,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6705917923849045e-05,
+      "loss": 3.4867,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.669754835514226e-05,
+      "loss": 3.4921,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.668916240763174e-05,
+      "loss": 3.4908,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6680776460121214e-05,
+      "loss": 3.4909,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6672390512610694e-05,
+      "loss": 3.4987,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.666402094390391e-05,
+      "loss": 3.4911,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6655634996393386e-05,
+      "loss": 3.4875,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6647249048882866e-05,
+      "loss": 3.4955,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6638863101372346e-05,
+      "loss": 3.4987,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6630493532665562e-05,
+      "loss": 3.4827,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.662210758515504e-05,
+      "loss": 3.4892,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.661372163764452e-05,
+      "loss": 3.4923,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6605335690134e-05,
+      "loss": 3.4867,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6596966121427214e-05,
+      "loss": 3.4813,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6588580173916688e-05,
+      "loss": 3.4861,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6580194226406168e-05,
+      "loss": 3.4892,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6571808278895648e-05,
+      "loss": 3.4912,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.656343871018886e-05,
+      "loss": 3.4846,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.655505276267834e-05,
+      "loss": 3.489,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.654666681516782e-05,
+      "loss": 3.5062,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.65382808676573e-05,
+      "loss": 3.5047,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.652991129895051e-05,
+      "loss": 3.4943,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6521525351439992e-05,
+      "loss": 3.4955,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6513139403929472e-05,
+      "loss": 3.4887,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6504753456418952e-05,
+      "loss": 3.488,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.649638388771216e-05,
+      "loss": 3.4814,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.648799794020164e-05,
+      "loss": 3.4828,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.647961199269112e-05,
+      "loss": 3.4895,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.64712260451806e-05,
+      "loss": 3.4984,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6462856476473814e-05,
+      "loss": 3.4898,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6454470528963294e-05,
+      "loss": 3.4913,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6446084581452774e-05,
+      "loss": 3.4901,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6437698633942254e-05,
+      "loss": 3.4838,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6429329065235463e-05,
+      "loss": 3.487,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6420943117724946e-05,
+      "loss": 3.5019,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6412557170214426e-05,
+      "loss": 3.5051,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6404171222703906e-05,
+      "loss": 3.4951,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6395801653997115e-05,
+      "loss": 3.4894,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6387415706486595e-05,
+      "loss": 3.4928,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6379029758976075e-05,
+      "loss": 3.494,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6370643811465555e-05,
+      "loss": 3.5033,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6362274242758767e-05,
+      "loss": 3.4896,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6353888295248247e-05,
+      "loss": 3.493,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6345502347737727e-05,
+      "loss": 3.4884,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6337132779030936e-05,
+      "loss": 3.491,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6328746831520416e-05,
+      "loss": 3.4918,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.63203608840099e-05,
+      "loss": 3.4917,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.631197493649938e-05,
+      "loss": 3.4966,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.630360536779259e-05,
+      "loss": 3.4828,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.629521942028207e-05,
+      "loss": 3.498,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.628683347277155e-05,
+      "loss": 3.4988,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.627844752526103e-05,
+      "loss": 3.4928,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.627007795655424e-05,
+      "loss": 3.4933,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.626169200904372e-05,
+      "loss": 3.4819,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.845421075820923,
+      "eval_runtime": 304.6112,
+      "eval_samples_per_second": 1252.715,
+      "eval_steps_per_second": 39.148,
+      "step": 1450080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.62533060615332e-05,
+      "loss": 3.4802,
+      "step": 1450496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.624492011402268e-05,
+      "loss": 3.4828,
+      "step": 1451008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.623653416651216e-05,
+      "loss": 3.5027,
+      "step": 1451520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.622814821900164e-05,
+      "loss": 3.4891,
+      "step": 1452032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6219778650294853e-05,
+      "loss": 3.5057,
+      "step": 1452544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6211392702784333e-05,
+      "loss": 3.4831,
+      "step": 1453056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6203006755273813e-05,
+      "loss": 3.4902,
+      "step": 1453568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6194620807763293e-05,
+      "loss": 3.4756,
+      "step": 1454080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6186251239056502e-05,
+      "loss": 3.4933,
+      "step": 1454592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6177865291545982e-05,
+      "loss": 3.4813,
+      "step": 1455104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6169479344035462e-05,
+      "loss": 3.5023,
+      "step": 1455616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6161109775328675e-05,
+      "loss": 3.4945,
+      "step": 1456128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6152723827818155e-05,
+      "loss": 3.4752,
+      "step": 1456640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6144337880307635e-05,
+      "loss": 3.4802,
+      "step": 1457152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6135951932797115e-05,
+      "loss": 3.4884,
+      "step": 1457664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6127582364090324e-05,
+      "loss": 3.4754,
+      "step": 1458176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6119196416579804e-05,
+      "loss": 3.485,
+      "step": 1458688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6110810469069287e-05,
+      "loss": 3.4874,
+      "step": 1459200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6102424521558767e-05,
+      "loss": 3.4812,
+      "step": 1459712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6094054952851976e-05,
+      "loss": 3.5079,
+      "step": 1460224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6085669005341456e-05,
+      "loss": 3.4896,
+      "step": 1460736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6077283057830936e-05,
+      "loss": 3.4894,
+      "step": 1461248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6068897110320416e-05,
+      "loss": 3.4882,
+      "step": 1461760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6060511162809896e-05,
+      "loss": 3.492,
+      "step": 1462272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.605212521529938e-05,
+      "loss": 3.4786,
+      "step": 1462784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.604375564659259e-05,
+      "loss": 3.4871,
+      "step": 1463296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6035369699082068e-05,
+      "loss": 3.4842,
+      "step": 1463808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6026983751571548e-05,
+      "loss": 3.4856,
+      "step": 1464320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6018597804061028e-05,
+      "loss": 3.4765,
+      "step": 1464832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6010211856550508e-05,
+      "loss": 3.4867,
+      "step": 1465344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6001825909039988e-05,
+      "loss": 3.4862,
+      "step": 1465856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5993439961529465e-05,
+      "loss": 3.4894,
+      "step": 1466368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5985054014018945e-05,
+      "loss": 3.4892,
+      "step": 1466880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.597668444531216e-05,
+      "loss": 3.4958,
+      "step": 1467392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.596829849780164e-05,
+      "loss": 3.4819,
+      "step": 1467904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5959912550291117e-05,
+      "loss": 3.49,
+      "step": 1468416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5951526602780597e-05,
+      "loss": 3.4763,
+      "step": 1468928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5943157034073813e-05,
+      "loss": 3.4909,
+      "step": 1469440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5934771086563293e-05,
+      "loss": 3.4728,
+      "step": 1469952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5926385139052766e-05,
+      "loss": 3.481,
+      "step": 1470464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5917999191542246e-05,
+      "loss": 3.479,
+      "step": 1470976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5909629622835462e-05,
+      "loss": 3.4962,
+      "step": 1471488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.590124367532494e-05,
+      "loss": 3.4784,
+      "step": 1472000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.589285772781442e-05,
+      "loss": 3.4836,
+      "step": 1472512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5884471780303898e-05,
+      "loss": 3.4945,
+      "step": 1473024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5876102211597114e-05,
+      "loss": 3.4907,
+      "step": 1473536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5867716264086587e-05,
+      "loss": 3.4906,
+      "step": 1474048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.585933031657607e-05,
+      "loss": 3.4729,
+      "step": 1474560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.585094436906555e-05,
+      "loss": 3.4735,
+      "step": 1475072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5842574800358767e-05,
+      "loss": 3.4881,
+      "step": 1475584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5834205231651976e-05,
+      "loss": 3.4883,
+      "step": 1476096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5825819284141456e-05,
+      "loss": 3.4776,
+      "step": 1476608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5817433336630936e-05,
+      "loss": 3.4724,
+      "step": 1477120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5809047389120412e-05,
+      "loss": 3.4849,
+      "step": 1477632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5800661441609892e-05,
+      "loss": 3.467,
+      "step": 1478144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5792275494099372e-05,
+      "loss": 3.487,
+      "step": 1478656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5783889546588852e-05,
+      "loss": 3.4719,
+      "step": 1479168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.577551997788206e-05,
+      "loss": 3.482,
+      "step": 1479680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.576713403037154e-05,
+      "loss": 3.4882,
+      "step": 1480192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5758748082861024e-05,
+      "loss": 3.4714,
+      "step": 1480704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5750362135350504e-05,
+      "loss": 3.4861,
+      "step": 1481216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5741992566643713e-05,
+      "loss": 3.4785,
+      "step": 1481728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5733606619133193e-05,
+      "loss": 3.4773,
+      "step": 1482240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5725220671622673e-05,
+      "loss": 3.4701,
+      "step": 1482752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5716834724112153e-05,
+      "loss": 3.4851,
+      "step": 1483264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5708465155405366e-05,
+      "loss": 3.4763,
+      "step": 1483776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5700079207894846e-05,
+      "loss": 3.4805,
+      "step": 1484288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5691693260384326e-05,
+      "loss": 3.4693,
+      "step": 1484800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5683307312873806e-05,
+      "loss": 3.4749,
+      "step": 1485312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5674937744167015e-05,
+      "loss": 3.4831,
+      "step": 1485824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5666551796656495e-05,
+      "loss": 3.4907,
+      "step": 1486336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5658182227949707e-05,
+      "loss": 3.4777,
+      "step": 1486848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5649796280439187e-05,
+      "loss": 3.4912,
+      "step": 1487360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5641410332928667e-05,
+      "loss": 3.4873,
+      "step": 1487872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5633024385418147e-05,
+      "loss": 3.4821,
+      "step": 1488384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5624638437907627e-05,
+      "loss": 3.4852,
+      "step": 1488896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5616252490397107e-05,
+      "loss": 3.4801,
+      "step": 1489408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5607866542886587e-05,
+      "loss": 3.4815,
+      "step": 1489920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.55994969741798e-05,
+      "loss": 3.4791,
+      "step": 1490432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.559111102666928e-05,
+      "loss": 3.4859,
+      "step": 1490944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.558272507915876e-05,
+      "loss": 3.4868,
+      "step": 1491456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.557433913164824e-05,
+      "loss": 3.4836,
+      "step": 1491968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.556595318413772e-05,
+      "loss": 3.4806,
+      "step": 1492480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.55575672366272e-05,
+      "loss": 3.4626,
+      "step": 1492992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.554918128911668e-05,
+      "loss": 3.4692,
+      "step": 1493504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.554079534160616e-05,
+      "loss": 3.478,
+      "step": 1494016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.553242577289937e-05,
+      "loss": 3.4805,
+      "step": 1494528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.552403982538885e-05,
+      "loss": 3.4742,
+      "step": 1495040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.551565387787833e-05,
+      "loss": 3.4827,
+      "step": 1495552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.550726793036781e-05,
+      "loss": 3.4777,
+      "step": 1496064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.549888198285729e-05,
+      "loss": 3.4707,
+      "step": 1496576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.549049603534677e-05,
+      "loss": 3.4776,
+      "step": 1497088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5482110087836248e-05,
+      "loss": 3.4857,
+      "step": 1497600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5473724140325728e-05,
+      "loss": 3.4836,
+      "step": 1498112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5465354571618944e-05,
+      "loss": 3.4913,
+      "step": 1498624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5456968624108424e-05,
+      "loss": 3.476,
+      "step": 1499136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5448582676597897e-05,
+      "loss": 3.4763,
+      "step": 1499648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.544019672908738e-05,
+      "loss": 3.4788,
+      "step": 1500160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5431827160380596e-05,
+      "loss": 3.4783,
+      "step": 1500672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.542344121287007e-05,
+      "loss": 3.4874,
+      "step": 1501184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.541505526535955e-05,
+      "loss": 3.4804,
+      "step": 1501696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.540666931784903e-05,
+      "loss": 3.477,
+      "step": 1502208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5398299749142245e-05,
+      "loss": 3.4819,
+      "step": 1502720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.538991380163172e-05,
+      "loss": 3.4882,
+      "step": 1503232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.53815278541212e-05,
+      "loss": 3.4712,
+      "step": 1503744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.537314190661068e-05,
+      "loss": 3.48,
+      "step": 1504256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5364772337903897e-05,
+      "loss": 3.4791,
+      "step": 1504768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.535638639039337e-05,
+      "loss": 3.474,
+      "step": 1505280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.534800044288285e-05,
+      "loss": 3.4675,
+      "step": 1505792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5339614495372334e-05,
+      "loss": 3.4766,
+      "step": 1506304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5331244926665543e-05,
+      "loss": 3.4778,
+      "step": 1506816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5322858979155023e-05,
+      "loss": 3.4772,
+      "step": 1507328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5314473031644503e-05,
+      "loss": 3.4742,
+      "step": 1507840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5306087084133983e-05,
+      "loss": 3.4749,
+      "step": 1508352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5297717515427195e-05,
+      "loss": 3.4936,
+      "step": 1508864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5289331567916675e-05,
+      "loss": 3.4905,
+      "step": 1509376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5280945620406155e-05,
+      "loss": 3.4839,
+      "step": 1509888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5272559672895635e-05,
+      "loss": 3.48,
+      "step": 1510400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5264190104188844e-05,
+      "loss": 3.4764,
+      "step": 1510912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5255804156678324e-05,
+      "loss": 3.4737,
+      "step": 1511424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5247418209167804e-05,
+      "loss": 3.4711,
+      "step": 1511936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5239032261657288e-05,
+      "loss": 3.4724,
+      "step": 1512448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5230662692950497e-05,
+      "loss": 3.4779,
+      "step": 1512960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5222276745439977e-05,
+      "loss": 3.4853,
+      "step": 1513472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5213890797929457e-05,
+      "loss": 3.4753,
+      "step": 1513984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5205521229222666e-05,
+      "loss": 3.4837,
+      "step": 1514496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.519713528171215e-05,
+      "loss": 3.4751,
+      "step": 1515008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.518874933420163e-05,
+      "loss": 3.4743,
+      "step": 1515520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.518036338669111e-05,
+      "loss": 3.472,
+      "step": 1516032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5171993817984318e-05,
+      "loss": 3.4909,
+      "step": 1516544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5163607870473798e-05,
+      "loss": 3.4954,
+      "step": 1517056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5155221922963278e-05,
+      "loss": 3.4822,
+      "step": 1517568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5146835975452758e-05,
+      "loss": 3.4767,
+      "step": 1518080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.513846640674597e-05,
+      "loss": 3.4823,
+      "step": 1518592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.513008045923545e-05,
+      "loss": 3.4789,
+      "step": 1519104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.512169451172493e-05,
+      "loss": 3.4897,
+      "step": 1519616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.511330856421441e-05,
+      "loss": 3.4734,
+      "step": 1520128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.510493899550762e-05,
+      "loss": 3.4832,
+      "step": 1520640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5096553047997103e-05,
+      "loss": 3.4756,
+      "step": 1521152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5088167100486583e-05,
+      "loss": 3.4845,
+      "step": 1521664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5079781152976063e-05,
+      "loss": 3.4765,
+      "step": 1522176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.507141158426927e-05,
+      "loss": 3.4776,
+      "step": 1522688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.506302563675875e-05,
+      "loss": 3.4853,
+      "step": 1523200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.505463968924823e-05,
+      "loss": 3.4739,
+      "step": 1523712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.504625374173771e-05,
+      "loss": 3.4838,
+      "step": 1524224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5037884173030924e-05,
+      "loss": 3.4892,
+      "step": 1524736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5029498225520404e-05,
+      "loss": 3.4776,
+      "step": 1525248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5021112278009884e-05,
+      "loss": 3.4788,
+      "step": 1525760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5012726330499364e-05,
+      "loss": 3.4756,
+      "step": 1526272
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8473012447357178,
+      "eval_runtime": 313.0038,
+      "eval_samples_per_second": 1219.126,
+      "eval_steps_per_second": 38.099,
+      "step": 1526400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5004356761792573e-05,
+      "loss": 3.4737,
+      "step": 1526784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4995970814282056e-05,
+      "loss": 3.4702,
+      "step": 1527296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4987584866771536e-05,
+      "loss": 3.4863,
+      "step": 1527808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4979198919261016e-05,
+      "loss": 3.4803,
+      "step": 1528320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.497082935055423e-05,
+      "loss": 3.492,
+      "step": 1528832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4962443403043705e-05,
+      "loss": 3.4721,
+      "step": 1529344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4954057455533185e-05,
+      "loss": 3.4797,
+      "step": 1529856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4945671508022665e-05,
+      "loss": 3.4597,
+      "step": 1530368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4937301939315878e-05,
+      "loss": 3.4843,
+      "step": 1530880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4928915991805358e-05,
+      "loss": 3.4685,
+      "step": 1531392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4920530044294838e-05,
+      "loss": 3.4857,
+      "step": 1531904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4912144096784318e-05,
+      "loss": 3.4852,
+      "step": 1532416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4903774528077527e-05,
+      "loss": 3.4619,
+      "step": 1532928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.489538858056701e-05,
+      "loss": 3.4734,
+      "step": 1533440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.488700263305649e-05,
+      "loss": 3.4736,
+      "step": 1533952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.487861668554597e-05,
+      "loss": 3.4639,
+      "step": 1534464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.487024711683918e-05,
+      "loss": 3.4738,
+      "step": 1534976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.486186116932866e-05,
+      "loss": 3.4766,
+      "step": 1535488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.485347522181814e-05,
+      "loss": 3.4746,
+      "step": 1536000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.484510565311135e-05,
+      "loss": 3.4925,
+      "step": 1536512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.483671970560083e-05,
+      "loss": 3.4741,
+      "step": 1537024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.482833375809031e-05,
+      "loss": 3.4758,
+      "step": 1537536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.481994781057979e-05,
+      "loss": 3.4781,
+      "step": 1538048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.481156186306927e-05,
+      "loss": 3.4799,
+      "step": 1538560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.480317591555875e-05,
+      "loss": 3.4695,
+      "step": 1539072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4794806346851964e-05,
+      "loss": 3.4819,
+      "step": 1539584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4786420399341444e-05,
+      "loss": 3.4683,
+      "step": 1540096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4778034451830924e-05,
+      "loss": 3.4703,
+      "step": 1540608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4769648504320404e-05,
+      "loss": 3.4632,
+      "step": 1541120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4761262556809883e-05,
+      "loss": 3.4761,
+      "step": 1541632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.475287660929936e-05,
+      "loss": 3.4749,
+      "step": 1542144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.474449066178884e-05,
+      "loss": 3.4772,
+      "step": 1542656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.473610471427832e-05,
+      "loss": 3.4771,
+      "step": 1543168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4727735145571532e-05,
+      "loss": 3.4846,
+      "step": 1543680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4719349198061012e-05,
+      "loss": 3.469,
+      "step": 1544192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4710963250550492e-05,
+      "loss": 3.4788,
+      "step": 1544704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4702577303039972e-05,
+      "loss": 3.4686,
+      "step": 1545216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.469420773433318e-05,
+      "loss": 3.473,
+      "step": 1545728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4685821786822665e-05,
+      "loss": 3.4624,
+      "step": 1546240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4677435839312145e-05,
+      "loss": 3.4681,
+      "step": 1546752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4669049891801625e-05,
+      "loss": 3.4681,
+      "step": 1547264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4660680323094834e-05,
+      "loss": 3.4831,
+      "step": 1547776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4652294375584314e-05,
+      "loss": 3.4659,
+      "step": 1548288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4643924806877526e-05,
+      "loss": 3.4706,
+      "step": 1548800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4635538859367006e-05,
+      "loss": 3.4848,
+      "step": 1549312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4627152911856486e-05,
+      "loss": 3.478,
+      "step": 1549824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4618766964345966e-05,
+      "loss": 3.4774,
+      "step": 1550336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4610381016835446e-05,
+      "loss": 3.4648,
+      "step": 1550848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4602011448128655e-05,
+      "loss": 3.4551,
+      "step": 1551360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4593625500618135e-05,
+      "loss": 3.4801,
+      "step": 1551872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.458523955310762e-05,
+      "loss": 3.474,
+      "step": 1552384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.45768536055971e-05,
+      "loss": 3.4626,
+      "step": 1552896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.456846765808658e-05,
+      "loss": 3.4661,
+      "step": 1553408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4560081710576058e-05,
+      "loss": 3.4738,
+      "step": 1553920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4551695763065535e-05,
+      "loss": 3.4544,
+      "step": 1554432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4543309815555015e-05,
+      "loss": 3.4743,
+      "step": 1554944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4534940246848227e-05,
+      "loss": 3.4606,
+      "step": 1555456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.452655429933771e-05,
+      "loss": 3.4716,
+      "step": 1555968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4518168351827187e-05,
+      "loss": 3.4735,
+      "step": 1556480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4509782404316667e-05,
+      "loss": 3.4639,
+      "step": 1556992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.450141283560988e-05,
+      "loss": 3.4687,
+      "step": 1557504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.449302688809936e-05,
+      "loss": 3.467,
+      "step": 1558016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4484640940588836e-05,
+      "loss": 3.4625,
+      "step": 1558528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.447625499307832e-05,
+      "loss": 3.4641,
+      "step": 1559040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4467885424371532e-05,
+      "loss": 3.4688,
+      "step": 1559552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.445949947686101e-05,
+      "loss": 3.466,
+      "step": 1560064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.445111352935049e-05,
+      "loss": 3.467,
+      "step": 1560576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.444272758183997e-05,
+      "loss": 3.4584,
+      "step": 1561088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.443435801313318e-05,
+      "loss": 3.4597,
+      "step": 1561600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.442597206562266e-05,
+      "loss": 3.4697,
+      "step": 1562112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.441758611811214e-05,
+      "loss": 3.4812,
+      "step": 1562624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4409216549405353e-05,
+      "loss": 3.4647,
+      "step": 1563136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4400830601894833e-05,
+      "loss": 3.4812,
+      "step": 1563648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.439244465438431e-05,
+      "loss": 3.4768,
+      "step": 1564160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.438405870687379e-05,
+      "loss": 3.4683,
+      "step": 1564672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4375672759363273e-05,
+      "loss": 3.4699,
+      "step": 1565184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4367286811852753e-05,
+      "loss": 3.477,
+      "step": 1565696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4358900864342233e-05,
+      "loss": 3.4682,
+      "step": 1566208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4350531295635442e-05,
+      "loss": 3.4684,
+      "step": 1566720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4342145348124922e-05,
+      "loss": 3.4754,
+      "step": 1567232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4333759400614402e-05,
+      "loss": 3.4711,
+      "step": 1567744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4325373453103882e-05,
+      "loss": 3.4734,
+      "step": 1568256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4316987505593365e-05,
+      "loss": 3.4687,
+      "step": 1568768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4308601558082842e-05,
+      "loss": 3.4533,
+      "step": 1569280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4300215610572322e-05,
+      "loss": 3.4558,
+      "step": 1569792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4291829663061802e-05,
+      "loss": 3.4677,
+      "step": 1570304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4283460094355014e-05,
+      "loss": 3.4697,
+      "step": 1570816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.427507414684449e-05,
+      "loss": 3.463,
+      "step": 1571328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4266688199333974e-05,
+      "loss": 3.4681,
+      "step": 1571840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4258302251823454e-05,
+      "loss": 3.4645,
+      "step": 1572352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4249932683116663e-05,
+      "loss": 3.4619,
+      "step": 1572864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4241546735606143e-05,
+      "loss": 3.4669,
+      "step": 1573376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4233160788095623e-05,
+      "loss": 3.4689,
+      "step": 1573888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4224774840585103e-05,
+      "loss": 3.4756,
+      "step": 1574400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4216405271878316e-05,
+      "loss": 3.4759,
+      "step": 1574912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4208019324367796e-05,
+      "loss": 3.4712,
+      "step": 1575424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4199633376857276e-05,
+      "loss": 3.4605,
+      "step": 1575936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4191247429346756e-05,
+      "loss": 3.4732,
+      "step": 1576448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4182877860639965e-05,
+      "loss": 3.4645,
+      "step": 1576960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4174491913129445e-05,
+      "loss": 3.4756,
+      "step": 1577472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4166105965618928e-05,
+      "loss": 3.4691,
+      "step": 1577984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4157720018108408e-05,
+      "loss": 3.4596,
+      "step": 1578496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4149350449401617e-05,
+      "loss": 3.4762,
+      "step": 1579008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4140964501891097e-05,
+      "loss": 3.472,
+      "step": 1579520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4132578554380577e-05,
+      "loss": 3.4643,
+      "step": 1580032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4124192606870057e-05,
+      "loss": 3.4681,
+      "step": 1580544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.411582303816327e-05,
+      "loss": 3.4688,
+      "step": 1581056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.410743709065275e-05,
+      "loss": 3.4543,
+      "step": 1581568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.409905114314223e-05,
+      "loss": 3.4577,
+      "step": 1582080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.409066519563171e-05,
+      "loss": 3.4691,
+      "step": 1582592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4082295626924918e-05,
+      "loss": 3.4654,
+      "step": 1583104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4073909679414398e-05,
+      "loss": 3.4647,
+      "step": 1583616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.406552373190388e-05,
+      "loss": 3.463,
+      "step": 1584128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.405713778439336e-05,
+      "loss": 3.4615,
+      "step": 1584640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.404876821568657e-05,
+      "loss": 3.4852,
+      "step": 1585152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.404038226817605e-05,
+      "loss": 3.4805,
+      "step": 1585664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.403199632066553e-05,
+      "loss": 3.4717,
+      "step": 1586176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.402361037315501e-05,
+      "loss": 3.4692,
+      "step": 1586688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4015240804448223e-05,
+      "loss": 3.4638,
+      "step": 1587200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4006854856937703e-05,
+      "loss": 3.4609,
+      "step": 1587712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3998468909427183e-05,
+      "loss": 3.4612,
+      "step": 1588224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3990082961916663e-05,
+      "loss": 3.4642,
+      "step": 1588736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3981713393209872e-05,
+      "loss": 3.4655,
+      "step": 1589248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3973327445699352e-05,
+      "loss": 3.4713,
+      "step": 1589760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3964941498188835e-05,
+      "loss": 3.4655,
+      "step": 1590272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3956555550678315e-05,
+      "loss": 3.4674,
+      "step": 1590784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3948185981971524e-05,
+      "loss": 3.4651,
+      "step": 1591296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3939800034461004e-05,
+      "loss": 3.4621,
+      "step": 1591808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3931414086950484e-05,
+      "loss": 3.4647,
+      "step": 1592320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3923028139439964e-05,
+      "loss": 3.474,
+      "step": 1592832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3914658570733177e-05,
+      "loss": 3.4838,
+      "step": 1593344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3906272623222657e-05,
+      "loss": 3.4741,
+      "step": 1593856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3897886675712137e-05,
+      "loss": 3.4628,
+      "step": 1594368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3889517107005346e-05,
+      "loss": 3.4721,
+      "step": 1594880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3881131159494826e-05,
+      "loss": 3.4662,
+      "step": 1595392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3872745211984306e-05,
+      "loss": 3.4771,
+      "step": 1595904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.386435926447379e-05,
+      "loss": 3.4659,
+      "step": 1596416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3855989695766998e-05,
+      "loss": 3.4674,
+      "step": 1596928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3847603748256478e-05,
+      "loss": 3.4648,
+      "step": 1597440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3839217800745958e-05,
+      "loss": 3.4757,
+      "step": 1597952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3830831853235438e-05,
+      "loss": 3.4628,
+      "step": 1598464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.382246228452865e-05,
+      "loss": 3.4669,
+      "step": 1598976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.381407633701813e-05,
+      "loss": 3.4716,
+      "step": 1599488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.380569038950761e-05,
+      "loss": 3.4642,
+      "step": 1600000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3797304441997087e-05,
+      "loss": 3.4714,
+      "step": 1600512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.37889348732903e-05,
+      "loss": 3.478,
+      "step": 1601024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.378054892577978e-05,
+      "loss": 3.4696,
+      "step": 1601536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.377216297826926e-05,
+      "loss": 3.4639,
+      "step": 1602048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.376377703075874e-05,
+      "loss": 3.464,
+      "step": 1602560
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.848076820373535,
+      "eval_runtime": 311.5948,
+      "eval_samples_per_second": 1224.638,
+      "eval_steps_per_second": 38.271,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.375539108324822e-05,
+      "loss": 3.4599,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.37470051357377e-05,
+      "loss": 3.4564,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.373861918822718e-05,
+      "loss": 3.4753,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.373023324071666e-05,
+      "loss": 3.4705,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3721863672009868e-05,
+      "loss": 3.4787,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.371347772449935e-05,
+      "loss": 3.4637,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.370509177698883e-05,
+      "loss": 3.4655,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.369670582947831e-05,
+      "loss": 3.4561,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.368833626077152e-05,
+      "loss": 3.4661,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3679950313261e-05,
+      "loss": 3.4625,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.367156436575048e-05,
+      "loss": 3.4707,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.366317841823996e-05,
+      "loss": 3.4744,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3654808849533173e-05,
+      "loss": 3.448,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3646422902022653e-05,
+      "loss": 3.464,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3638036954512133e-05,
+      "loss": 3.4638,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3629651007001613e-05,
+      "loss": 3.4536,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3621281438294822e-05,
+      "loss": 3.4632,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3612895490784305e-05,
+      "loss": 3.4627,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3604509543273785e-05,
+      "loss": 3.4611,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3596139974566994e-05,
+      "loss": 3.4784,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3587754027056474e-05,
+      "loss": 3.4626,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3579368079545954e-05,
+      "loss": 3.4714,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3570982132035434e-05,
+      "loss": 3.465,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3562596184524914e-05,
+      "loss": 3.4662,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3554210237014394e-05,
+      "loss": 3.4575,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3545840668307606e-05,
+      "loss": 3.4686,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3537454720797086e-05,
+      "loss": 3.4628,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3529068773286566e-05,
+      "loss": 3.4576,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3520682825776043e-05,
+      "loss": 3.4542,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3512296878265523e-05,
+      "loss": 3.464,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3503910930755006e-05,
+      "loss": 3.4633,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3495524983244486e-05,
+      "loss": 3.467,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3487139035733966e-05,
+      "loss": 3.4686,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3478769467027175e-05,
+      "loss": 3.472,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3470383519516655e-05,
+      "loss": 3.4579,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3461997572006135e-05,
+      "loss": 3.466,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3453611624495615e-05,
+      "loss": 3.4599,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3445242055788828e-05,
+      "loss": 3.4612,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3436856108278308e-05,
+      "loss": 3.4491,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3428470160767787e-05,
+      "loss": 3.4539,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3420084213257267e-05,
+      "loss": 3.4594,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3411731023354212e-05,
+      "loss": 3.4695,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.340334507584369e-05,
+      "loss": 3.4567,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.339495912833317e-05,
+      "loss": 3.4594,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.338657318082265e-05,
+      "loss": 3.4763,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.337818723331213e-05,
+      "loss": 3.4662,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.336980128580161e-05,
+      "loss": 3.4676,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.336141533829109e-05,
+      "loss": 3.4536,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.335302939078057e-05,
+      "loss": 3.4429,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.334465982207378e-05,
+      "loss": 3.4668,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.333627387456326e-05,
+      "loss": 3.4646,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.332788792705274e-05,
+      "loss": 3.4512,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.331950197954222e-05,
+      "loss": 3.4556,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.331113241083543e-05,
+      "loss": 3.4647,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3302746463324914e-05,
+      "loss": 3.4388,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3294360515814394e-05,
+      "loss": 3.465,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.328597456830387e-05,
+      "loss": 3.4537,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3277604999597083e-05,
+      "loss": 3.4556,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3269219052086563e-05,
+      "loss": 3.465,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3260833104576042e-05,
+      "loss": 3.4566,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3252447157065522e-05,
+      "loss": 3.4544,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3244077588358735e-05,
+      "loss": 3.4598,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3235691640848215e-05,
+      "loss": 3.453,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.322730569333769e-05,
+      "loss": 3.4471,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.321891974582717e-05,
+      "loss": 3.4583,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3210550177120384e-05,
+      "loss": 3.4582,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3202164229609867e-05,
+      "loss": 3.4537,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3193778282099344e-05,
+      "loss": 3.4506,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3185392334588824e-05,
+      "loss": 3.4482,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3177022765882036e-05,
+      "loss": 3.4545,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3168636818371516e-05,
+      "loss": 3.4668,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.316026724966473e-05,
+      "loss": 3.4536,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.315188130215421e-05,
+      "loss": 3.4714,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.314349535464369e-05,
+      "loss": 3.4668,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3135109407133165e-05,
+      "loss": 3.4585,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3126723459622645e-05,
+      "loss": 3.4534,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3118337512112125e-05,
+      "loss": 3.4674,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3109951564601605e-05,
+      "loss": 3.4571,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3101565617091085e-05,
+      "loss": 3.4533,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3093196048384297e-05,
+      "loss": 3.4662,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3084810100873777e-05,
+      "loss": 3.4605,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3076424153363257e-05,
+      "loss": 3.465,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3068038205852737e-05,
+      "loss": 3.4582,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3059652258342217e-05,
+      "loss": 3.4451,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3051266310831697e-05,
+      "loss": 3.4397,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3042880363321177e-05,
+      "loss": 3.4545,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.303451079461439e-05,
+      "loss": 3.4623,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.302612484710387e-05,
+      "loss": 3.4528,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3017738899593346e-05,
+      "loss": 3.456,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3009352952082826e-05,
+      "loss": 3.4566,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.300098338337604e-05,
+      "loss": 3.4477,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2992597435865522e-05,
+      "loss": 3.4581,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2984211488355e-05,
+      "loss": 3.4532,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.297582554084448e-05,
+      "loss": 3.4652,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.296745597213769e-05,
+      "loss": 3.4664,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.295907002462717e-05,
+      "loss": 3.459,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.295068407711665e-05,
+      "loss": 3.451,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.294229812960613e-05,
+      "loss": 3.4565,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2933928560899343e-05,
+      "loss": 3.4545,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.292554261338882e-05,
+      "loss": 3.467,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.29171566658783e-05,
+      "loss": 3.4564,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.290877071836778e-05,
+      "loss": 3.45,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2900401149660992e-05,
+      "loss": 3.4617,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2892015202150472e-05,
+      "loss": 3.4608,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2883629254639952e-05,
+      "loss": 3.4539,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2875243307129432e-05,
+      "loss": 3.4574,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2866873738422645e-05,
+      "loss": 3.4539,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.285848779091212e-05,
+      "loss": 3.4506,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.28501018434016e-05,
+      "loss": 3.4395,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2841715895891085e-05,
+      "loss": 3.457,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2833346327184294e-05,
+      "loss": 3.4585,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2824960379673774e-05,
+      "loss": 3.4517,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2816574432163254e-05,
+      "loss": 3.4548,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2808188484652733e-05,
+      "loss": 3.4473,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.279983529474968e-05,
+      "loss": 3.4719,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.279144934723916e-05,
+      "loss": 3.4693,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.278306339972864e-05,
+      "loss": 3.4621,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.277467745221812e-05,
+      "loss": 3.4595,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2766291504707595e-05,
+      "loss": 3.4523,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2757905557197075e-05,
+      "loss": 3.45,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2749519609686555e-05,
+      "loss": 3.449,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2741133662176038e-05,
+      "loss": 3.4525,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2732764093469247e-05,
+      "loss": 3.4576,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2724378145958727e-05,
+      "loss": 3.4569,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2715992198448207e-05,
+      "loss": 3.4586,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2707606250937687e-05,
+      "loss": 3.4535,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.26992366822309e-05,
+      "loss": 3.4566,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.269085073472038e-05,
+      "loss": 3.4513,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268246478720986e-05,
+      "loss": 3.454,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.267407883969934e-05,
+      "loss": 3.4552,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.266570927099255e-05,
+      "loss": 3.4708,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.265732332348203e-05,
+      "loss": 3.4666,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.264893737597151e-05,
+      "loss": 3.4532,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2640551428460992e-05,
+      "loss": 3.4614,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.26321818597542e-05,
+      "loss": 3.4579,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262379591224368e-05,
+      "loss": 3.4611,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.261540996473316e-05,
+      "loss": 3.4553,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.260704039602637e-05,
+      "loss": 3.4612,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2598654448515853e-05,
+      "loss": 3.4535,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2590268501005333e-05,
+      "loss": 3.4617,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2581882553494813e-05,
+      "loss": 3.4504,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2573496605984293e-05,
+      "loss": 3.4581,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.256511065847377e-05,
+      "loss": 3.4566,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.255672471096325e-05,
+      "loss": 3.4578,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.254833876345273e-05,
+      "loss": 3.459,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2539969194745946e-05,
+      "loss": 3.4601,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2531583247235422e-05,
+      "loss": 3.4672,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2523197299724902e-05,
+      "loss": 3.4486,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2514811352214382e-05,
+      "loss": 3.4565,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8491785526275635,
+      "eval_runtime": 309.9768,
+      "eval_samples_per_second": 1231.031,
+      "eval_steps_per_second": 38.471,
+      "step": 1679040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2506441783507594e-05,
+      "loss": 3.4554,
+      "step": 1679360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2498055835997074e-05,
+      "loss": 3.4427,
+      "step": 1679872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2489669888486554e-05,
+      "loss": 3.4646,
+      "step": 1680384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2481283940976034e-05,
+      "loss": 3.4597,
+      "step": 1680896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2472914372269243e-05,
+      "loss": 3.4686,
+      "step": 1681408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2464528424758723e-05,
+      "loss": 3.4521,
+      "step": 1681920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2456142477248203e-05,
+      "loss": 3.4494,
+      "step": 1682432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2447772908541416e-05,
+      "loss": 3.448,
+      "step": 1682944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2439386961030896e-05,
+      "loss": 3.454,
+      "step": 1683456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2431001013520376e-05,
+      "loss": 3.4511,
+      "step": 1683968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2422615066009856e-05,
+      "loss": 3.4601,
+      "step": 1684480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2414245497303068e-05,
+      "loss": 3.4653,
+      "step": 1684992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2405859549792545e-05,
+      "loss": 3.4393,
+      "step": 1685504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2397473602282025e-05,
+      "loss": 3.4483,
+      "step": 1686016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2389087654771508e-05,
+      "loss": 3.456,
+      "step": 1686528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2380718086064717e-05,
+      "loss": 3.4402,
+      "step": 1687040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2372332138554197e-05,
+      "loss": 3.449,
+      "step": 1687552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2363946191043677e-05,
+      "loss": 3.4509,
+      "step": 1688064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2355560243533157e-05,
+      "loss": 3.4518,
+      "step": 1688576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.234719067482637e-05,
+      "loss": 3.4657,
+      "step": 1689088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.233880472731585e-05,
+      "loss": 3.4569,
+      "step": 1689600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.233041877980533e-05,
+      "loss": 3.4567,
+      "step": 1690112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.232203283229481e-05,
+      "loss": 3.4546,
+      "step": 1690624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.231366326358802e-05,
+      "loss": 3.4553,
+      "step": 1691136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.23052773160775e-05,
+      "loss": 3.446,
+      "step": 1691648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2296907747370714e-05,
+      "loss": 3.4575,
+      "step": 1692160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.228852179986019e-05,
+      "loss": 3.4489,
+      "step": 1692672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.228013585234967e-05,
+      "loss": 3.4455,
+      "step": 1693184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.227174990483915e-05,
+      "loss": 3.4441,
+      "step": 1693696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.226336395732863e-05,
+      "loss": 3.4514,
+      "step": 1694208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.225497800981811e-05,
+      "loss": 3.4529,
+      "step": 1694720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.224659206230759e-05,
+      "loss": 3.4529,
+      "step": 1695232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.223820611479707e-05,
+      "loss": 3.458,
+      "step": 1695744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2229836546090283e-05,
+      "loss": 3.4628,
+      "step": 1696256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2221450598579763e-05,
+      "loss": 3.4478,
+      "step": 1696768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2213064651069243e-05,
+      "loss": 3.4531,
+      "step": 1697280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2204678703558723e-05,
+      "loss": 3.4522,
+      "step": 1697792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2196309134851932e-05,
+      "loss": 3.4486,
+      "step": 1698304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2187923187341415e-05,
+      "loss": 3.4377,
+      "step": 1698816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2179537239830895e-05,
+      "loss": 3.4432,
+      "step": 1699328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2171151292320372e-05,
+      "loss": 3.448,
+      "step": 1699840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2162781723613584e-05,
+      "loss": 3.4572,
+      "step": 1700352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2154395776103064e-05,
+      "loss": 3.4481,
+      "step": 1700864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2146009828592544e-05,
+      "loss": 3.4462,
+      "step": 1701376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2137623881082024e-05,
+      "loss": 3.4665,
+      "step": 1701888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2129254312375237e-05,
+      "loss": 3.453,
+      "step": 1702400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2120868364864717e-05,
+      "loss": 3.4579,
+      "step": 1702912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2112482417354197e-05,
+      "loss": 3.4424,
+      "step": 1703424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2104096469843673e-05,
+      "loss": 3.4329,
+      "step": 1703936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2095726901136886e-05,
+      "loss": 3.4597,
+      "step": 1704448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.208734095362637e-05,
+      "loss": 3.4493,
+      "step": 1704960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2078955006115846e-05,
+      "loss": 3.4426,
+      "step": 1705472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2070569058605326e-05,
+      "loss": 3.4459,
+      "step": 1705984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2062215868702267e-05,
+      "loss": 3.4504,
+      "step": 1706496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2053829921191747e-05,
+      "loss": 3.4284,
+      "step": 1707008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.204544397368123e-05,
+      "loss": 3.4487,
+      "step": 1707520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.203705802617071e-05,
+      "loss": 3.4445,
+      "step": 1708032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.202867207866019e-05,
+      "loss": 3.4405,
+      "step": 1708544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.202028613114967e-05,
+      "loss": 3.457,
+      "step": 1709056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2011900183639147e-05,
+      "loss": 3.4424,
+      "step": 1709568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2003514236128627e-05,
+      "loss": 3.4432,
+      "step": 1710080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.199514466742184e-05,
+      "loss": 3.4513,
+      "step": 1710592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.198675871991132e-05,
+      "loss": 3.4415,
+      "step": 1711104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.19783727724008e-05,
+      "loss": 3.4404,
+      "step": 1711616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.196998682489028e-05,
+      "loss": 3.4454,
+      "step": 1712128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1961617256183492e-05,
+      "loss": 3.4453,
+      "step": 1712640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.195323130867297e-05,
+      "loss": 3.4443,
+      "step": 1713152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1944845361162448e-05,
+      "loss": 3.4409,
+      "step": 1713664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.193645941365193e-05,
+      "loss": 3.4324,
+      "step": 1714176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.192808984494514e-05,
+      "loss": 3.4449,
+      "step": 1714688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.191970389743462e-05,
+      "loss": 3.4555,
+      "step": 1715200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1911334328727833e-05,
+      "loss": 3.4396,
+      "step": 1715712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1902948381217313e-05,
+      "loss": 3.4625,
+      "step": 1716224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1894562433706793e-05,
+      "loss": 3.4535,
+      "step": 1716736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1886176486196273e-05,
+      "loss": 3.4537,
+      "step": 1717248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1877790538685753e-05,
+      "loss": 3.4378,
+      "step": 1717760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1869404591175233e-05,
+      "loss": 3.458,
+      "step": 1718272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1861018643664713e-05,
+      "loss": 3.4468,
+      "step": 1718784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1852632696154193e-05,
+      "loss": 3.4396,
+      "step": 1719296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1844263127447402e-05,
+      "loss": 3.4585,
+      "step": 1719808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1835877179936885e-05,
+      "loss": 3.449,
+      "step": 1720320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1827491232426365e-05,
+      "loss": 3.455,
+      "step": 1720832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1819105284915845e-05,
+      "loss": 3.4475,
+      "step": 1721344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1810719337405322e-05,
+      "loss": 3.4311,
+      "step": 1721856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1802333389894802e-05,
+      "loss": 3.4307,
+      "step": 1722368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.179394744238428e-05,
+      "loss": 3.4396,
+      "step": 1722880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1785577873677494e-05,
+      "loss": 3.4513,
+      "step": 1723392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1777191926166974e-05,
+      "loss": 3.4438,
+      "step": 1723904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1768805978656454e-05,
+      "loss": 3.4412,
+      "step": 1724416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1760420031145934e-05,
+      "loss": 3.4467,
+      "step": 1724928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1752050462439147e-05,
+      "loss": 3.4376,
+      "step": 1725440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1743664514928623e-05,
+      "loss": 3.443,
+      "step": 1725952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1735278567418106e-05,
+      "loss": 3.4409,
+      "step": 1726464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1726892619907586e-05,
+      "loss": 3.4592,
+      "step": 1726976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1718523051200795e-05,
+      "loss": 3.4557,
+      "step": 1727488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1710137103690275e-05,
+      "loss": 3.4476,
+      "step": 1728000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1701751156179755e-05,
+      "loss": 3.4422,
+      "step": 1728512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1693365208669235e-05,
+      "loss": 3.4476,
+      "step": 1729024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1684995639962448e-05,
+      "loss": 3.4382,
+      "step": 1729536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1676609692451928e-05,
+      "loss": 3.4577,
+      "step": 1730048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1668223744941408e-05,
+      "loss": 3.4459,
+      "step": 1730560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1659837797430888e-05,
+      "loss": 3.4428,
+      "step": 1731072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1651468228724097e-05,
+      "loss": 3.4475,
+      "step": 1731584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1643082281213577e-05,
+      "loss": 3.4523,
+      "step": 1732096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1634696333703057e-05,
+      "loss": 3.4391,
+      "step": 1732608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.162631038619254e-05,
+      "loss": 3.4484,
+      "step": 1733120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.161794081748575e-05,
+      "loss": 3.4387,
+      "step": 1733632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.160955486997523e-05,
+      "loss": 3.4378,
+      "step": 1734144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.160116892246471e-05,
+      "loss": 3.4307,
+      "step": 1734656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.159278297495419e-05,
+      "loss": 3.4453,
+      "step": 1735168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.15844134062474e-05,
+      "loss": 3.4492,
+      "step": 1735680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.157602745873688e-05,
+      "loss": 3.4394,
+      "step": 1736192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.156764151122636e-05,
+      "loss": 3.4455,
+      "step": 1736704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.155925556371584e-05,
+      "loss": 3.4402,
+      "step": 1737216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.155088599500905e-05,
+      "loss": 3.4573,
+      "step": 1737728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.154250004749853e-05,
+      "loss": 3.459,
+      "step": 1738240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.153411409998801e-05,
+      "loss": 3.4547,
+      "step": 1738752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1525728152477494e-05,
+      "loss": 3.4482,
+      "step": 1739264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1517358583770703e-05,
+      "loss": 3.4407,
+      "step": 1739776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1508972636260183e-05,
+      "loss": 3.4401,
+      "step": 1740288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1500586688749663e-05,
+      "loss": 3.4345,
+      "step": 1740800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1492200741239143e-05,
+      "loss": 3.4413,
+      "step": 1741312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1483831172532355e-05,
+      "loss": 3.4467,
+      "step": 1741824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1475445225021835e-05,
+      "loss": 3.443,
+      "step": 1742336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1467059277511315e-05,
+      "loss": 3.4475,
+      "step": 1742848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1458673330000795e-05,
+      "loss": 3.4453,
+      "step": 1743360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1450303761294004e-05,
+      "loss": 3.4484,
+      "step": 1743872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1441917813783484e-05,
+      "loss": 3.4398,
+      "step": 1744384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1433531866272964e-05,
+      "loss": 3.4461,
+      "step": 1744896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1425145918762447e-05,
+      "loss": 3.4433,
+      "step": 1745408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1416776350055656e-05,
+      "loss": 3.4555,
+      "step": 1745920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1408390402545136e-05,
+      "loss": 3.4582,
+      "step": 1746432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1400004455034616e-05,
+      "loss": 3.4406,
+      "step": 1746944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1391618507524096e-05,
+      "loss": 3.4522,
+      "step": 1747456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.138324893881731e-05,
+      "loss": 3.4441,
+      "step": 1747968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.137486299130679e-05,
+      "loss": 3.451,
+      "step": 1748480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.136647704379627e-05,
+      "loss": 3.443,
+      "step": 1748992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1358091096285745e-05,
+      "loss": 3.4514,
+      "step": 1749504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1349721527578958e-05,
+      "loss": 3.4393,
+      "step": 1750016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1341335580068438e-05,
+      "loss": 3.4547,
+      "step": 1750528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1332949632557918e-05,
+      "loss": 3.4363,
+      "step": 1751040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.132458006385113e-05,
+      "loss": 3.4467,
+      "step": 1751552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.131619411634061e-05,
+      "loss": 3.4458,
+      "step": 1752064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.130780816883009e-05,
+      "loss": 3.4481,
+      "step": 1752576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.129942222131957e-05,
+      "loss": 3.4446,
+      "step": 1753088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.129105265261278e-05,
+      "loss": 3.4509,
+      "step": 1753600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1282666705102262e-05,
+      "loss": 3.4531,
+      "step": 1754112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1274280757591742e-05,
+      "loss": 3.4368,
+      "step": 1754624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.126589481008122e-05,
+      "loss": 3.4517,
+      "step": 1755136
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.850489854812622,
+      "eval_runtime": 313.3866,
+      "eval_samples_per_second": 1217.637,
+      "eval_steps_per_second": 38.052,
+      "step": 1755360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.12575088625707e-05,
+      "loss": 3.4407,
+      "step": 1755648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.124912291506018e-05,
+      "loss": 3.4353,
+      "step": 1756160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.124073696754966e-05,
+      "loss": 3.4492,
+      "step": 1756672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.123235102003914e-05,
+      "loss": 3.4501,
+      "step": 1757184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.122398145133235e-05,
+      "loss": 3.4555,
+      "step": 1757696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.121559550382183e-05,
+      "loss": 3.443,
+      "step": 1758208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.120720955631131e-05,
+      "loss": 3.44,
+      "step": 1758720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.119882360880079e-05,
+      "loss": 3.4404,
+      "step": 1759232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1190454040094e-05,
+      "loss": 3.4405,
+      "step": 1759744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.118206809258348e-05,
+      "loss": 3.4378,
+      "step": 1760256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1173682145072964e-05,
+      "loss": 3.4484,
+      "step": 1760768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1165296197562444e-05,
+      "loss": 3.4515,
+      "step": 1761280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1156926628855653e-05,
+      "loss": 3.4364,
+      "step": 1761792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1148540681345133e-05,
+      "loss": 3.4355,
+      "step": 1762304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1140154733834613e-05,
+      "loss": 3.4435,
+      "step": 1762816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1131768786324093e-05,
+      "loss": 3.427,
+      "step": 1763328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1123399217617305e-05,
+      "loss": 3.4407,
+      "step": 1763840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1115013270106785e-05,
+      "loss": 3.4387,
+      "step": 1764352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1106627322596265e-05,
+      "loss": 3.4431,
+      "step": 1764864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1098241375085745e-05,
+      "loss": 3.4508,
+      "step": 1765376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1089871806378954e-05,
+      "loss": 3.4489,
+      "step": 1765888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1081485858868434e-05,
+      "loss": 3.4441,
+      "step": 1766400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1073099911357917e-05,
+      "loss": 3.4436,
+      "step": 1766912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1064713963847397e-05,
+      "loss": 3.4425,
+      "step": 1767424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1056328016336874e-05,
+      "loss": 3.4364,
+      "step": 1767936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1047942068826354e-05,
+      "loss": 3.4465,
+      "step": 1768448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1039572500119566e-05,
+      "loss": 3.4424,
+      "step": 1768960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1031186552609046e-05,
+      "loss": 3.434,
+      "step": 1769472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1022800605098526e-05,
+      "loss": 3.4343,
+      "step": 1769984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1014414657588006e-05,
+      "loss": 3.4459,
+      "step": 1770496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1006028710077486e-05,
+      "loss": 3.4379,
+      "step": 1771008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0997642762566966e-05,
+      "loss": 3.4435,
+      "step": 1771520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0989256815056446e-05,
+      "loss": 3.4487,
+      "step": 1772032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0980887246349655e-05,
+      "loss": 3.449,
+      "step": 1772544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.097250129883914e-05,
+      "loss": 3.4339,
+      "step": 1773056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.096411535132862e-05,
+      "loss": 3.4436,
+      "step": 1773568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0955729403818098e-05,
+      "loss": 3.4431,
+      "step": 1774080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0947359835111307e-05,
+      "loss": 3.4357,
+      "step": 1774592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0938973887600787e-05,
+      "loss": 3.4289,
+      "step": 1775104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0930587940090267e-05,
+      "loss": 3.4292,
+      "step": 1775616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0922201992579747e-05,
+      "loss": 3.4392,
+      "step": 1776128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.091383242387296e-05,
+      "loss": 3.4458,
+      "step": 1776640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.090544647636244e-05,
+      "loss": 3.4343,
+      "step": 1777152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.089706052885192e-05,
+      "loss": 3.435,
+      "step": 1777664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.08886745813414e-05,
+      "loss": 3.4532,
+      "step": 1778176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.088030501263461e-05,
+      "loss": 3.444,
+      "step": 1778688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.087191906512409e-05,
+      "loss": 3.447,
+      "step": 1779200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0863533117613572e-05,
+      "loss": 3.4356,
+      "step": 1779712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0855147170103052e-05,
+      "loss": 3.4168,
+      "step": 1780224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.084677760139626e-05,
+      "loss": 3.45,
+      "step": 1780736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.083839165388574e-05,
+      "loss": 3.4385,
+      "step": 1781248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.083000570637522e-05,
+      "loss": 3.4324,
+      "step": 1781760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.08216197588647e-05,
+      "loss": 3.4388,
+      "step": 1782272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0813266568961646e-05,
+      "loss": 3.4353,
+      "step": 1782784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0804880621451123e-05,
+      "loss": 3.4256,
+      "step": 1783296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0796494673940602e-05,
+      "loss": 3.4314,
+      "step": 1783808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0788108726430082e-05,
+      "loss": 3.4313,
+      "step": 1784320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0779722778919562e-05,
+      "loss": 3.4319,
+      "step": 1784832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0771336831409042e-05,
+      "loss": 3.4408,
+      "step": 1785344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0762950883898526e-05,
+      "loss": 3.4345,
+      "step": 1785856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0754564936388002e-05,
+      "loss": 3.4331,
+      "step": 1786368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0746195367681215e-05,
+      "loss": 3.4448,
+      "step": 1786880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0737809420170695e-05,
+      "loss": 3.4257,
+      "step": 1787392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0729423472660175e-05,
+      "loss": 3.4286,
+      "step": 1787904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0721037525149655e-05,
+      "loss": 3.4313,
+      "step": 1788416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0712667956442867e-05,
+      "loss": 3.4362,
+      "step": 1788928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0704282008932347e-05,
+      "loss": 3.4294,
+      "step": 1789440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0695896061421824e-05,
+      "loss": 3.4329,
+      "step": 1789952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0687510113911304e-05,
+      "loss": 3.418,
+      "step": 1790464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0679140545204516e-05,
+      "loss": 3.4352,
+      "step": 1790976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0670754597693996e-05,
+      "loss": 3.4458,
+      "step": 1791488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.066238502898721e-05,
+      "loss": 3.4284,
+      "step": 1792000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.065399908147669e-05,
+      "loss": 3.4511,
+      "step": 1792512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.064561313396617e-05,
+      "loss": 3.4413,
+      "step": 1793024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.063722718645565e-05,
+      "loss": 3.4467,
+      "step": 1793536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0628841238945125e-05,
+      "loss": 3.4282,
+      "step": 1794048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0620455291434608e-05,
+      "loss": 3.4459,
+      "step": 1794560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0612069343924088e-05,
+      "loss": 3.4354,
+      "step": 1795072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0603683396413568e-05,
+      "loss": 3.4312,
+      "step": 1795584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0595313827706777e-05,
+      "loss": 3.4482,
+      "step": 1796096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0586927880196257e-05,
+      "loss": 3.4361,
+      "step": 1796608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0578541932685737e-05,
+      "loss": 3.4435,
+      "step": 1797120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0570155985175217e-05,
+      "loss": 3.4377,
+      "step": 1797632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0561770037664697e-05,
+      "loss": 3.4173,
+      "step": 1798144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.055338409015418e-05,
+      "loss": 3.4265,
+      "step": 1798656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0544998142643657e-05,
+      "loss": 3.4249,
+      "step": 1799168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0536612195133137e-05,
+      "loss": 3.4396,
+      "step": 1799680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.052824262642635e-05,
+      "loss": 3.4356,
+      "step": 1800192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.051985667891583e-05,
+      "loss": 3.4298,
+      "step": 1800704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.051147073140531e-05,
+      "loss": 3.4351,
+      "step": 1801216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0503101162698522e-05,
+      "loss": 3.4301,
+      "step": 1801728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0494715215188002e-05,
+      "loss": 3.4321,
+      "step": 1802240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.048632926767748e-05,
+      "loss": 3.4314,
+      "step": 1802752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.047794332016696e-05,
+      "loss": 3.4474,
+      "step": 1803264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.046957375146017e-05,
+      "loss": 3.4511,
+      "step": 1803776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.046118780394965e-05,
+      "loss": 3.432,
+      "step": 1804288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.045280185643913e-05,
+      "loss": 3.4306,
+      "step": 1804800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.044441590892861e-05,
+      "loss": 3.4364,
+      "step": 1805312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0436046340221823e-05,
+      "loss": 3.4281,
+      "step": 1805824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0427660392711303e-05,
+      "loss": 3.4452,
+      "step": 1806336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.041927444520078e-05,
+      "loss": 3.4347,
+      "step": 1806848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0410888497690263e-05,
+      "loss": 3.4336,
+      "step": 1807360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0402518928983476e-05,
+      "loss": 3.4385,
+      "step": 1807872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0394132981472952e-05,
+      "loss": 3.4393,
+      "step": 1808384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0385747033962432e-05,
+      "loss": 3.4271,
+      "step": 1808896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0377361086451912e-05,
+      "loss": 3.4344,
+      "step": 1809408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0368991517745124e-05,
+      "loss": 3.4332,
+      "step": 1809920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0360605570234604e-05,
+      "loss": 3.4255,
+      "step": 1810432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0352219622724084e-05,
+      "loss": 3.4214,
+      "step": 1810944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0343833675213564e-05,
+      "loss": 3.4334,
+      "step": 1811456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0335464106506777e-05,
+      "loss": 3.4372,
+      "step": 1811968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0327078158996253e-05,
+      "loss": 3.4314,
+      "step": 1812480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0318692211485733e-05,
+      "loss": 3.4348,
+      "step": 1812992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0310306263975217e-05,
+      "loss": 3.4276,
+      "step": 1813504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0301936695268426e-05,
+      "loss": 3.4478,
+      "step": 1814016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0293550747757906e-05,
+      "loss": 3.4471,
+      "step": 1814528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0285164800247386e-05,
+      "loss": 3.4433,
+      "step": 1815040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0276778852736866e-05,
+      "loss": 3.437,
+      "step": 1815552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0268409284030078e-05,
+      "loss": 3.4249,
+      "step": 1816064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0260023336519558e-05,
+      "loss": 3.4337,
+      "step": 1816576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0251637389009038e-05,
+      "loss": 3.4248,
+      "step": 1817088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0243251441498518e-05,
+      "loss": 3.4334,
+      "step": 1817600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0234881872791727e-05,
+      "loss": 3.432,
+      "step": 1818112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0226495925281207e-05,
+      "loss": 3.4355,
+      "step": 1818624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0218109977770687e-05,
+      "loss": 3.4329,
+      "step": 1819136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.020972403026017e-05,
+      "loss": 3.4337,
+      "step": 1819648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.020135446155338e-05,
+      "loss": 3.4375,
+      "step": 1820160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.019296851404286e-05,
+      "loss": 3.4306,
+      "step": 1820672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.018458256653234e-05,
+      "loss": 3.4352,
+      "step": 1821184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.017619661902182e-05,
+      "loss": 3.4345,
+      "step": 1821696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0167827050315032e-05,
+      "loss": 3.4406,
+      "step": 1822208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0159441102804512e-05,
+      "loss": 3.4488,
+      "step": 1822720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0151055155293992e-05,
+      "loss": 3.4306,
+      "step": 1823232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.014266920778347e-05,
+      "loss": 3.4428,
+      "step": 1823744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.013429963907668e-05,
+      "loss": 3.4345,
+      "step": 1824256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.012591369156616e-05,
+      "loss": 3.4395,
+      "step": 1824768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.011752774405564e-05,
+      "loss": 3.4334,
+      "step": 1825280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.010914179654512e-05,
+      "loss": 3.4379,
+      "step": 1825792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0100772227838333e-05,
+      "loss": 3.4319,
+      "step": 1826304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0092386280327813e-05,
+      "loss": 3.4435,
+      "step": 1826816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0084000332817293e-05,
+      "loss": 3.4257,
+      "step": 1827328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0075630764110502e-05,
+      "loss": 3.4376,
+      "step": 1827840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0067244816599985e-05,
+      "loss": 3.4366,
+      "step": 1828352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0058858869089465e-05,
+      "loss": 3.4397,
+      "step": 1828864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0050472921578945e-05,
+      "loss": 3.433,
+      "step": 1829376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0042103352872154e-05,
+      "loss": 3.4371,
+      "step": 1829888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0033717405361634e-05,
+      "loss": 3.4459,
+      "step": 1830400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0025331457851114e-05,
+      "loss": 3.4264,
+      "step": 1830912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0016945510340594e-05,
+      "loss": 3.442,
+      "step": 1831424
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8520917892456055,
+      "eval_runtime": 302.8687,
+      "eval_samples_per_second": 1259.922,
+      "eval_steps_per_second": 39.373,
+      "step": 1831680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0008575941633807e-05,
+      "loss": 3.4273,
+      "step": 1831936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0000189994123287e-05,
+      "loss": 3.4313,
+      "step": 1832448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9991804046612767e-05,
+      "loss": 3.4348,
+      "step": 1832960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9983418099102247e-05,
+      "loss": 3.4377,
+      "step": 1833472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9975048530395456e-05,
+      "loss": 3.4477,
+      "step": 1833984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9966662582884936e-05,
+      "loss": 3.4292,
+      "step": 1834496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.995827663537442e-05,
+      "loss": 3.4331,
+      "step": 1835008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.99498906878639e-05,
+      "loss": 3.4308,
+      "step": 1835520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9941521119157108e-05,
+      "loss": 3.4272,
+      "step": 1836032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9933135171646588e-05,
+      "loss": 3.4284,
+      "step": 1836544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9924749224136068e-05,
+      "loss": 3.4387,
+      "step": 1837056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9916363276625548e-05,
+      "loss": 3.4397,
+      "step": 1837568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.990799370791876e-05,
+      "loss": 3.4234,
+      "step": 1838080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.989960776040824e-05,
+      "loss": 3.4229,
+      "step": 1838592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.989122181289772e-05,
+      "loss": 3.4357,
+      "step": 1839104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.98828358653872e-05,
+      "loss": 3.4162,
+      "step": 1839616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.987446629668041e-05,
+      "loss": 3.4301,
+      "step": 1840128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.986608034916989e-05,
+      "loss": 3.4274,
+      "step": 1840640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9857694401659373e-05,
+      "loss": 3.4346,
+      "step": 1841152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.984930845414885e-05,
+      "loss": 3.4355,
+      "step": 1841664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9840938885442062e-05,
+      "loss": 3.4413,
+      "step": 1842176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9832552937931542e-05,
+      "loss": 3.4366,
+      "step": 1842688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9824166990421022e-05,
+      "loss": 3.4316,
+      "step": 1843200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.98157810429105e-05,
+      "loss": 3.4269,
+      "step": 1843712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.980739509539998e-05,
+      "loss": 3.4281,
+      "step": 1844224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.979900914788946e-05,
+      "loss": 3.4371,
+      "step": 1844736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.979062320037894e-05,
+      "loss": 3.4298,
+      "step": 1845248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.978225363167215e-05,
+      "loss": 3.4232,
+      "step": 1845760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.977386768416163e-05,
+      "loss": 3.427,
+      "step": 1846272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.976548173665111e-05,
+      "loss": 3.4314,
+      "step": 1846784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9757095789140594e-05,
+      "loss": 3.4268,
+      "step": 1847296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9748726220433803e-05,
+      "loss": 3.4387,
+      "step": 1847808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9740340272923283e-05,
+      "loss": 3.4348,
+      "step": 1848320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9731954325412763e-05,
+      "loss": 3.4357,
+      "step": 1848832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9723568377902243e-05,
+      "loss": 3.425,
+      "step": 1849344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9715198809195455e-05,
+      "loss": 3.4365,
+      "step": 1849856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9706812861684935e-05,
+      "loss": 3.4323,
+      "step": 1850368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9698426914174415e-05,
+      "loss": 3.4291,
+      "step": 1850880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9690040966663895e-05,
+      "loss": 3.4169,
+      "step": 1851392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9681671397957104e-05,
+      "loss": 3.4203,
+      "step": 1851904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9673285450446584e-05,
+      "loss": 3.4226,
+      "step": 1852416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9664899502936064e-05,
+      "loss": 3.4374,
+      "step": 1852928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9656513555425544e-05,
+      "loss": 3.4267,
+      "step": 1853440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9648143986718757e-05,
+      "loss": 3.424,
+      "step": 1853952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9639758039208237e-05,
+      "loss": 3.4409,
+      "step": 1854464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9631372091697717e-05,
+      "loss": 3.4333,
+      "step": 1854976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9622986144187197e-05,
+      "loss": 3.4403,
+      "step": 1855488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9614616575480406e-05,
+      "loss": 3.423,
+      "step": 1856000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.960623062796989e-05,
+      "loss": 3.4052,
+      "step": 1856512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.959784468045937e-05,
+      "loss": 3.4425,
+      "step": 1857024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.958945873294885e-05,
+      "loss": 3.4317,
+      "step": 1857536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9581089164242058e-05,
+      "loss": 3.4237,
+      "step": 1858048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9572703216731538e-05,
+      "loss": 3.4253,
+      "step": 1858560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.956433364802475e-05,
+      "loss": 3.4271,
+      "step": 1859072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.955594770051423e-05,
+      "loss": 3.4151,
+      "step": 1859584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.954756175300371e-05,
+      "loss": 3.4188,
+      "step": 1860096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.953917580549319e-05,
+      "loss": 3.4243,
+      "step": 1860608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.953078985798267e-05,
+      "loss": 3.4207,
+      "step": 1861120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.952240391047215e-05,
+      "loss": 3.435,
+      "step": 1861632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9514017962961627e-05,
+      "loss": 3.422,
+      "step": 1862144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.950563201545111e-05,
+      "loss": 3.4219,
+      "step": 1862656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9497262446744323e-05,
+      "loss": 3.4369,
+      "step": 1863168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9488876499233803e-05,
+      "loss": 3.4126,
+      "step": 1863680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.948049055172328e-05,
+      "loss": 3.4159,
+      "step": 1864192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.947210460421276e-05,
+      "loss": 3.4246,
+      "step": 1864704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.946373503550597e-05,
+      "loss": 3.4238,
+      "step": 1865216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.945534908799545e-05,
+      "loss": 3.4209,
+      "step": 1865728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.944696314048493e-05,
+      "loss": 3.4254,
+      "step": 1866240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.943857719297441e-05,
+      "loss": 3.4094,
+      "step": 1866752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9430207624267624e-05,
+      "loss": 3.4205,
+      "step": 1867264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.94218216767571e-05,
+      "loss": 3.434,
+      "step": 1867776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9413452108050313e-05,
+      "loss": 3.4186,
+      "step": 1868288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9405066160539796e-05,
+      "loss": 3.4394,
+      "step": 1868800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9396680213029273e-05,
+      "loss": 3.4299,
+      "step": 1869312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9388294265518753e-05,
+      "loss": 3.438,
+      "step": 1869824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9379908318008233e-05,
+      "loss": 3.416,
+      "step": 1870336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9371522370497713e-05,
+      "loss": 3.4332,
+      "step": 1870848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9363136422987193e-05,
+      "loss": 3.4285,
+      "step": 1871360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9354750475476673e-05,
+      "loss": 3.4172,
+      "step": 1871872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9346380906769885e-05,
+      "loss": 3.437,
+      "step": 1872384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9337994959259365e-05,
+      "loss": 3.4263,
+      "step": 1872896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9329609011748845e-05,
+      "loss": 3.4303,
+      "step": 1873408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9321223064238325e-05,
+      "loss": 3.4286,
+      "step": 1873920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9312837116727805e-05,
+      "loss": 3.41,
+      "step": 1874432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.930445116921728e-05,
+      "loss": 3.4132,
+      "step": 1874944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9296065221706765e-05,
+      "loss": 3.417,
+      "step": 1875456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9287679274196245e-05,
+      "loss": 3.4266,
+      "step": 1875968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9279309705489454e-05,
+      "loss": 3.4247,
+      "step": 1876480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9270923757978934e-05,
+      "loss": 3.419,
+      "step": 1876992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9262537810468414e-05,
+      "loss": 3.424,
+      "step": 1877504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9254151862957894e-05,
+      "loss": 3.4222,
+      "step": 1878016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9245765915447374e-05,
+      "loss": 3.4198,
+      "step": 1878528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9237379967936857e-05,
+      "loss": 3.4211,
+      "step": 1879040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9228994020426337e-05,
+      "loss": 3.439,
+      "step": 1879552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9220624451719546e-05,
+      "loss": 3.4359,
+      "step": 1880064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9212238504209026e-05,
+      "loss": 3.4267,
+      "step": 1880576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9203852556698506e-05,
+      "loss": 3.4162,
+      "step": 1881088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9195466609187986e-05,
+      "loss": 3.4251,
+      "step": 1881600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.91870970404812e-05,
+      "loss": 3.417,
+      "step": 1882112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.917871109297068e-05,
+      "loss": 3.4338,
+      "step": 1882624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.917032514546016e-05,
+      "loss": 3.4263,
+      "step": 1883136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9161939197949635e-05,
+      "loss": 3.4222,
+      "step": 1883648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9153569629242847e-05,
+      "loss": 3.4266,
+      "step": 1884160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9145183681732327e-05,
+      "loss": 3.4326,
+      "step": 1884672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.913679773422181e-05,
+      "loss": 3.4175,
+      "step": 1885184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9128411786711287e-05,
+      "loss": 3.4245,
+      "step": 1885696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.91200422180045e-05,
+      "loss": 3.4213,
+      "step": 1886208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.911165627049398e-05,
+      "loss": 3.418,
+      "step": 1886720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.910327032298346e-05,
+      "loss": 3.4119,
+      "step": 1887232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9094884375472936e-05,
+      "loss": 3.4224,
+      "step": 1887744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9086514806766152e-05,
+      "loss": 3.4239,
+      "step": 1888256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9078128859255632e-05,
+      "loss": 3.4247,
+      "step": 1888768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.906974291174511e-05,
+      "loss": 3.4237,
+      "step": 1889280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.906135696423459e-05,
+      "loss": 3.4167,
+      "step": 1889792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.90529873955278e-05,
+      "loss": 3.4349,
+      "step": 1890304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.904460144801728e-05,
+      "loss": 3.4401,
+      "step": 1890816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.903621550050676e-05,
+      "loss": 3.4299,
+      "step": 1891328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.902782955299624e-05,
+      "loss": 3.4327,
+      "step": 1891840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9019459984289453e-05,
+      "loss": 3.4094,
+      "step": 1892352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9011074036778933e-05,
+      "loss": 3.4252,
+      "step": 1892864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.900268808926841e-05,
+      "loss": 3.4186,
+      "step": 1893376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.899430214175789e-05,
+      "loss": 3.4201,
+      "step": 1893888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8985932573051106e-05,
+      "loss": 3.4217,
+      "step": 1894400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8977546625540582e-05,
+      "loss": 3.4213,
+      "step": 1894912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8969160678030062e-05,
+      "loss": 3.4245,
+      "step": 1895424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8960774730519542e-05,
+      "loss": 3.4237,
+      "step": 1895936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8952405161812755e-05,
+      "loss": 3.4286,
+      "step": 1896448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8944019214302235e-05,
+      "loss": 3.4221,
+      "step": 1896960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8935633266791715e-05,
+      "loss": 3.4219,
+      "step": 1897472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8927247319281195e-05,
+      "loss": 3.4255,
+      "step": 1897984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8918877750574407e-05,
+      "loss": 3.4334,
+      "step": 1898496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8910491803063884e-05,
+      "loss": 3.4383,
+      "step": 1899008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8902105855553364e-05,
+      "loss": 3.4199,
+      "step": 1899520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8893719908042844e-05,
+      "loss": 3.4328,
+      "step": 1900032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8885350339336056e-05,
+      "loss": 3.4242,
+      "step": 1900544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8876964391825536e-05,
+      "loss": 3.4274,
+      "step": 1901056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8868578444315016e-05,
+      "loss": 3.4247,
+      "step": 1901568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8860192496804496e-05,
+      "loss": 3.4255,
+      "step": 1902080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8851822928097705e-05,
+      "loss": 3.4214,
+      "step": 1902592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.884343698058719e-05,
+      "loss": 3.4336,
+      "step": 1903104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.883505103307667e-05,
+      "loss": 3.4166,
+      "step": 1903616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.882668146436988e-05,
+      "loss": 3.426,
+      "step": 1904128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8818295516859357e-05,
+      "loss": 3.4271,
+      "step": 1904640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8809909569348837e-05,
+      "loss": 3.4322,
+      "step": 1905152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8801523621838317e-05,
+      "loss": 3.4189,
+      "step": 1905664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.879315405313153e-05,
+      "loss": 3.4263,
+      "step": 1906176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.878476810562101e-05,
+      "loss": 3.4326,
+      "step": 1906688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.877638215811049e-05,
+      "loss": 3.4224,
+      "step": 1907200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.876799621059997e-05,
+      "loss": 3.4276,
+      "step": 1907712
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.853714942932129,
+      "eval_runtime": 303.083,
+      "eval_samples_per_second": 1259.031,
+      "eval_steps_per_second": 39.346,
+      "step": 1908000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.875961026308945e-05,
+      "loss": 3.406,
+      "step": 1908224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.875122431557893e-05,
+      "loss": 3.4204,
+      "step": 1908736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.874283836806841e-05,
+      "loss": 3.4254,
+      "step": 1909248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8734468799361622e-05,
+      "loss": 3.4296,
+      "step": 1909760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8726082851851102e-05,
+      "loss": 3.4378,
+      "step": 1910272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8717696904340582e-05,
+      "loss": 3.4184,
+      "step": 1910784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.870931095683006e-05,
+      "loss": 3.4227,
+      "step": 1911296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.870092500931954e-05,
+      "loss": 3.4225,
+      "step": 1911808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.869253906180902e-05,
+      "loss": 3.4139,
+      "step": 1912320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.86841531142985e-05,
+      "loss": 3.42,
+      "step": 1912832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.867578354559171e-05,
+      "loss": 3.4279,
+      "step": 1913344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.866739759808119e-05,
+      "loss": 3.426,
+      "step": 1913856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.865901165057067e-05,
+      "loss": 3.4231,
+      "step": 1914368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8650642081863883e-05,
+      "loss": 3.4085,
+      "step": 1914880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.864225613435336e-05,
+      "loss": 3.4253,
+      "step": 1915392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8633870186842843e-05,
+      "loss": 3.4056,
+      "step": 1915904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8625484239332323e-05,
+      "loss": 3.4155,
+      "step": 1916416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8617098291821803e-05,
+      "loss": 3.4197,
+      "step": 1916928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8608712344311283e-05,
+      "loss": 3.4209,
+      "step": 1917440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8600326396800763e-05,
+      "loss": 3.4251,
+      "step": 1917952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.859194044929024e-05,
+      "loss": 3.4318,
+      "step": 1918464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8583570880583452e-05,
+      "loss": 3.4271,
+      "step": 1918976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8575184933072935e-05,
+      "loss": 3.4241,
+      "step": 1919488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8566798985562415e-05,
+      "loss": 3.4122,
+      "step": 1920000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8558413038051892e-05,
+      "loss": 3.4175,
+      "step": 1920512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8550043469345104e-05,
+      "loss": 3.425,
+      "step": 1921024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8541673900638313e-05,
+      "loss": 3.4211,
+      "step": 1921536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8533287953127797e-05,
+      "loss": 3.416,
+      "step": 1922048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8524902005617277e-05,
+      "loss": 3.4135,
+      "step": 1922560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8516516058106757e-05,
+      "loss": 3.4214,
+      "step": 1923072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8508130110596237e-05,
+      "loss": 3.4117,
+      "step": 1923584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8499744163085713e-05,
+      "loss": 3.4284,
+      "step": 1924096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8491358215575193e-05,
+      "loss": 3.4233,
+      "step": 1924608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8482988646868406e-05,
+      "loss": 3.4256,
+      "step": 1925120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.847460269935789e-05,
+      "loss": 3.4176,
+      "step": 1925632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8466216751847366e-05,
+      "loss": 3.4212,
+      "step": 1926144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8457830804336846e-05,
+      "loss": 3.4258,
+      "step": 1926656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8449461235630058e-05,
+      "loss": 3.4156,
+      "step": 1927168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8441075288119538e-05,
+      "loss": 3.4104,
+      "step": 1927680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8432689340609015e-05,
+      "loss": 3.4092,
+      "step": 1928192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8424303393098498e-05,
+      "loss": 3.411,
+      "step": 1928704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.841595020319544e-05,
+      "loss": 3.4288,
+      "step": 1929216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.840756425568492e-05,
+      "loss": 3.415,
+      "step": 1929728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.83991783081744e-05,
+      "loss": 3.4163,
+      "step": 1930240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.839079236066388e-05,
+      "loss": 3.4269,
+      "step": 1930752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.838240641315336e-05,
+      "loss": 3.4225,
+      "step": 1931264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.837402046564284e-05,
+      "loss": 3.4272,
+      "step": 1931776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.836563451813232e-05,
+      "loss": 3.4164,
+      "step": 1932288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.83572485706218e-05,
+      "loss": 3.3975,
+      "step": 1932800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8348879001915012e-05,
+      "loss": 3.4283,
+      "step": 1933312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8340493054404488e-05,
+      "loss": 3.4166,
+      "step": 1933824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8332107106893968e-05,
+      "loss": 3.4169,
+      "step": 1934336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.832372115938345e-05,
+      "loss": 3.4159,
+      "step": 1934848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8315367969480393e-05,
+      "loss": 3.4155,
+      "step": 1935360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8306982021969873e-05,
+      "loss": 3.4021,
+      "step": 1935872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8298596074459353e-05,
+      "loss": 3.4112,
+      "step": 1936384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8290210126948833e-05,
+      "loss": 3.4133,
+      "step": 1936896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8281824179438313e-05,
+      "loss": 3.4099,
+      "step": 1937408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8273438231927793e-05,
+      "loss": 3.4282,
+      "step": 1937920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8265052284417273e-05,
+      "loss": 3.4098,
+      "step": 1938432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8256666336906753e-05,
+      "loss": 3.4106,
+      "step": 1938944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8248296768199962e-05,
+      "loss": 3.4272,
+      "step": 1939456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8239910820689442e-05,
+      "loss": 3.4018,
+      "step": 1939968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8231524873178922e-05,
+      "loss": 3.4037,
+      "step": 1940480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8223138925668405e-05,
+      "loss": 3.4178,
+      "step": 1940992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8214769356961614e-05,
+      "loss": 3.4136,
+      "step": 1941504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8206399788254827e-05,
+      "loss": 3.4095,
+      "step": 1942016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8198013840744307e-05,
+      "loss": 3.4186,
+      "step": 1942528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8189627893233783e-05,
+      "loss": 3.3992,
+      "step": 1943040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8181241945723267e-05,
+      "loss": 3.4082,
+      "step": 1943552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8172855998212747e-05,
+      "loss": 3.4262,
+      "step": 1944064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8164470050702227e-05,
+      "loss": 3.4101,
+      "step": 1944576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8156100481995436e-05,
+      "loss": 3.4257,
+      "step": 1945088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8147714534484916e-05,
+      "loss": 3.4193,
+      "step": 1945600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8139328586974396e-05,
+      "loss": 3.4327,
+      "step": 1946112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8130942639463876e-05,
+      "loss": 3.4067,
+      "step": 1946624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.812255669195336e-05,
+      "loss": 3.4221,
+      "step": 1947136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.811417074444284e-05,
+      "loss": 3.4182,
+      "step": 1947648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8105784796932315e-05,
+      "loss": 3.4046,
+      "step": 1948160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8097415228225528e-05,
+      "loss": 3.4298,
+      "step": 1948672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8089029280715008e-05,
+      "loss": 3.4145,
+      "step": 1949184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8080643333204488e-05,
+      "loss": 3.4195,
+      "step": 1949696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8072257385693968e-05,
+      "loss": 3.4196,
+      "step": 1950208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8063871438183448e-05,
+      "loss": 3.4025,
+      "step": 1950720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8055485490672928e-05,
+      "loss": 3.3992,
+      "step": 1951232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8047099543162408e-05,
+      "loss": 3.4084,
+      "step": 1951744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8038713595651888e-05,
+      "loss": 3.4157,
+      "step": 1952256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8030344026945097e-05,
+      "loss": 3.4156,
+      "step": 1952768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8021958079434577e-05,
+      "loss": 3.4077,
+      "step": 1953280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.801357213192406e-05,
+      "loss": 3.4108,
+      "step": 1953792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.800518618441354e-05,
+      "loss": 3.4177,
+      "step": 1954304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.799681661570675e-05,
+      "loss": 3.4097,
+      "step": 1954816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.798843066819623e-05,
+      "loss": 3.4056,
+      "step": 1955328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.798004472068571e-05,
+      "loss": 3.4303,
+      "step": 1955840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.797167515197892e-05,
+      "loss": 3.4249,
+      "step": 1956352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.79632892044684e-05,
+      "loss": 3.4189,
+      "step": 1956864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.795490325695788e-05,
+      "loss": 3.4058,
+      "step": 1957376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.794651730944736e-05,
+      "loss": 3.4185,
+      "step": 1957888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.793814774074057e-05,
+      "loss": 3.4086,
+      "step": 1958400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.792976179323005e-05,
+      "loss": 3.423,
+      "step": 1958912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.792137584571953e-05,
+      "loss": 3.4135,
+      "step": 1959424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7912989898209014e-05,
+      "loss": 3.4099,
+      "step": 1959936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7904620329502223e-05,
+      "loss": 3.4161,
+      "step": 1960448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7896234381991703e-05,
+      "loss": 3.4233,
+      "step": 1960960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7887848434481183e-05,
+      "loss": 3.4068,
+      "step": 1961472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7879462486970663e-05,
+      "loss": 3.4162,
+      "step": 1961984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7871092918263875e-05,
+      "loss": 3.4061,
+      "step": 1962496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7862706970753355e-05,
+      "loss": 3.4101,
+      "step": 1963008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7854321023242835e-05,
+      "loss": 3.4032,
+      "step": 1963520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7845935075732315e-05,
+      "loss": 3.4058,
+      "step": 1964032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7837565507025524e-05,
+      "loss": 3.4143,
+      "step": 1964544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7829179559515004e-05,
+      "loss": 3.4154,
+      "step": 1965056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7820793612004484e-05,
+      "loss": 3.414,
+      "step": 1965568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7812407664493967e-05,
+      "loss": 3.4069,
+      "step": 1966080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7804038095787176e-05,
+      "loss": 3.4255,
+      "step": 1966592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7795652148276656e-05,
+      "loss": 3.4263,
+      "step": 1967104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7787266200766136e-05,
+      "loss": 3.4168,
+      "step": 1967616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7778880253255616e-05,
+      "loss": 3.4245,
+      "step": 1968128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.777051068454883e-05,
+      "loss": 3.4037,
+      "step": 1968640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7762141115842038e-05,
+      "loss": 3.4119,
+      "step": 1969152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7753755168331518e-05,
+      "loss": 3.4104,
+      "step": 1969664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7745369220820998e-05,
+      "loss": 3.4065,
+      "step": 1970176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7736983273310478e-05,
+      "loss": 3.4093,
+      "step": 1970688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7728597325799958e-05,
+      "loss": 3.4122,
+      "step": 1971200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7720211378289438e-05,
+      "loss": 3.4146,
+      "step": 1971712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7711825430778918e-05,
+      "loss": 3.4133,
+      "step": 1972224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7703439483268398e-05,
+      "loss": 3.4166,
+      "step": 1972736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.769506991456161e-05,
+      "loss": 3.4141,
+      "step": 1973248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.768668396705109e-05,
+      "loss": 3.4081,
+      "step": 1973760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7678298019540567e-05,
+      "loss": 3.4139,
+      "step": 1974272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7669928450833782e-05,
+      "loss": 3.422,
+      "step": 1974784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7661542503323262e-05,
+      "loss": 3.4267,
+      "step": 1975296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.765315655581274e-05,
+      "loss": 3.41,
+      "step": 1975808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.764477060830222e-05,
+      "loss": 3.425,
+      "step": 1976320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.763640103959543e-05,
+      "loss": 3.4123,
+      "step": 1976832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.762801509208491e-05,
+      "loss": 3.4177,
+      "step": 1977344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.761962914457439e-05,
+      "loss": 3.416,
+      "step": 1977856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.761124319706387e-05,
+      "loss": 3.4114,
+      "step": 1978368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7602873628357084e-05,
+      "loss": 3.411,
+      "step": 1978880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7594487680846564e-05,
+      "loss": 3.4297,
+      "step": 1979392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.758610173333604e-05,
+      "loss": 3.4003,
+      "step": 1979904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.757771578582552e-05,
+      "loss": 3.4189,
+      "step": 1980416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7569346217118736e-05,
+      "loss": 3.4181,
+      "step": 1980928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7560960269608213e-05,
+      "loss": 3.4156,
+      "step": 1981440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7552574322097693e-05,
+      "loss": 3.4095,
+      "step": 1981952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7544188374587173e-05,
+      "loss": 3.416,
+      "step": 1982464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7535818805880385e-05,
+      "loss": 3.4209,
+      "step": 1982976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.752743285836986e-05,
+      "loss": 3.4164,
+      "step": 1983488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7519046910859345e-05,
+      "loss": 3.4171,
+      "step": 1984000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8546881675720215,
+      "eval_runtime": 302.2415,
+      "eval_samples_per_second": 1262.537,
+      "eval_steps_per_second": 39.455,
+      "step": 1984320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7510660963348825e-05,
+      "loss": 3.4039,
+      "step": 1984512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7502291394642034e-05,
+      "loss": 3.4091,
+      "step": 1985024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7493905447131514e-05,
+      "loss": 3.4139,
+      "step": 1985536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7485519499620994e-05,
+      "loss": 3.4175,
+      "step": 1986048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7477133552110474e-05,
+      "loss": 3.4291,
+      "step": 1986560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7468763983403686e-05,
+      "loss": 3.4093,
+      "step": 1987072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7460378035893166e-05,
+      "loss": 3.4138,
+      "step": 1987584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7451992088382646e-05,
+      "loss": 3.4119,
+      "step": 1988096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7443606140872126e-05,
+      "loss": 3.4056,
+      "step": 1988608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7435236572165335e-05,
+      "loss": 3.4076,
+      "step": 1989120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7426850624654815e-05,
+      "loss": 3.4165,
+      "step": 1989632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.74184646771443e-05,
+      "loss": 3.4183,
+      "step": 1990144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.741007872963378e-05,
+      "loss": 3.4171,
+      "step": 1990656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7401709160926988e-05,
+      "loss": 3.3943,
+      "step": 1991168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7393323213416468e-05,
+      "loss": 3.4162,
+      "step": 1991680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7384937265905948e-05,
+      "loss": 3.3953,
+      "step": 1992192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7376551318395428e-05,
+      "loss": 3.408,
+      "step": 1992704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.736818174968864e-05,
+      "loss": 3.4085,
+      "step": 1993216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.735979580217812e-05,
+      "loss": 3.4111,
+      "step": 1993728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.73514098546676e-05,
+      "loss": 3.4159,
+      "step": 1994240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.734304028596081e-05,
+      "loss": 3.4222,
+      "step": 1994752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.733465433845029e-05,
+      "loss": 3.4173,
+      "step": 1995264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.732626839093977e-05,
+      "loss": 3.4109,
+      "step": 1995776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7317882443429252e-05,
+      "loss": 3.4013,
+      "step": 1996288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7309496495918732e-05,
+      "loss": 3.4133,
+      "step": 1996800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7301110548408212e-05,
+      "loss": 3.4155,
+      "step": 1997312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.729274097970142e-05,
+      "loss": 3.4108,
+      "step": 1997824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.72843550321909e-05,
+      "loss": 3.4068,
+      "step": 1998336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.727596908468038e-05,
+      "loss": 3.4011,
+      "step": 1998848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.726758313716986e-05,
+      "loss": 3.4122,
+      "step": 1999360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.725919718965934e-05,
+      "loss": 3.4025,
+      "step": 1999872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.725081124214882e-05,
+      "loss": 3.4197,
+      "step": 2000384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.72424252946383e-05,
+      "loss": 3.4177,
+      "step": 2000896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.723403934712778e-05,
+      "loss": 3.4131,
+      "step": 2001408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.722566977842099e-05,
+      "loss": 3.4107,
+      "step": 2001920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.721728383091047e-05,
+      "loss": 3.4048,
+      "step": 2002432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7208897883399953e-05,
+      "loss": 3.4177,
+      "step": 2002944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7200528314693163e-05,
+      "loss": 3.4083,
+      "step": 2003456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7192142367182642e-05,
+      "loss": 3.3957,
+      "step": 2003968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7183756419672122e-05,
+      "loss": 3.4032,
+      "step": 2004480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7175370472161602e-05,
+      "loss": 3.4035,
+      "step": 2004992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7167000903454815e-05,
+      "loss": 3.4124,
+      "step": 2005504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7158614955944295e-05,
+      "loss": 3.4062,
+      "step": 2006016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7150229008433775e-05,
+      "loss": 3.406,
+      "step": 2006528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7141843060923255e-05,
+      "loss": 3.4166,
+      "step": 2007040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7133473492216464e-05,
+      "loss": 3.4134,
+      "step": 2007552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7125087544705944e-05,
+      "loss": 3.4142,
+      "step": 2008064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7116701597195424e-05,
+      "loss": 3.4106,
+      "step": 2008576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7108315649684907e-05,
+      "loss": 3.3834,
+      "step": 2009088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7099946080978116e-05,
+      "loss": 3.4183,
+      "step": 2009600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7091560133467596e-05,
+      "loss": 3.4083,
+      "step": 2010112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7083174185957076e-05,
+      "loss": 3.4058,
+      "step": 2010624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7074788238446556e-05,
+      "loss": 3.4076,
+      "step": 2011136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.70664350485435e-05,
+      "loss": 3.4049,
+      "step": 2011648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.705804910103298e-05,
+      "loss": 3.392,
+      "step": 2012160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.704966315352246e-05,
+      "loss": 3.3975,
+      "step": 2012672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7041277206011938e-05,
+      "loss": 3.4043,
+      "step": 2013184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7032891258501418e-05,
+      "loss": 3.3996,
+      "step": 2013696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7024505310990897e-05,
+      "loss": 3.4141,
+      "step": 2014208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7016119363480377e-05,
+      "loss": 3.4009,
+      "step": 2014720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.700773341596986e-05,
+      "loss": 3.4014,
+      "step": 2015232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.699936384726307e-05,
+      "loss": 3.4156,
+      "step": 2015744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.699097789975255e-05,
+      "loss": 3.3958,
+      "step": 2016256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.698259195224203e-05,
+      "loss": 3.3913,
+      "step": 2016768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.697420600473151e-05,
+      "loss": 3.409,
+      "step": 2017280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6965836436024722e-05,
+      "loss": 3.4004,
+      "step": 2017792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6957450488514202e-05,
+      "loss": 3.3975,
+      "step": 2018304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6949064541003682e-05,
+      "loss": 3.4059,
+      "step": 2018816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6940678593493162e-05,
+      "loss": 3.3932,
+      "step": 2019328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.693230902478637e-05,
+      "loss": 3.3999,
+      "step": 2019840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.692392307727585e-05,
+      "loss": 3.4158,
+      "step": 2020352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.691553712976533e-05,
+      "loss": 3.401,
+      "step": 2020864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6907167561058544e-05,
+      "loss": 3.4147,
+      "step": 2021376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6898781613548024e-05,
+      "loss": 3.4062,
+      "step": 2021888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6890395666037503e-05,
+      "loss": 3.4238,
+      "step": 2022400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6882009718526983e-05,
+      "loss": 3.3972,
+      "step": 2022912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6873623771016463e-05,
+      "loss": 3.4131,
+      "step": 2023424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.686523782350594e-05,
+      "loss": 3.4091,
+      "step": 2023936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6856851875995423e-05,
+      "loss": 3.3929,
+      "step": 2024448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6848482307288636e-05,
+      "loss": 3.4173,
+      "step": 2024960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6840096359778112e-05,
+      "loss": 3.4061,
+      "step": 2025472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6831710412267592e-05,
+      "loss": 3.4105,
+      "step": 2025984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6823324464757072e-05,
+      "loss": 3.4092,
+      "step": 2026496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6814938517246552e-05,
+      "loss": 3.393,
+      "step": 2027008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6806552569736032e-05,
+      "loss": 3.3893,
+      "step": 2027520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6798166622225516e-05,
+      "loss": 3.3953,
+      "step": 2028032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6789780674714995e-05,
+      "loss": 3.4081,
+      "step": 2028544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6781411106008205e-05,
+      "loss": 3.4001,
+      "step": 2029056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6773025158497685e-05,
+      "loss": 3.3991,
+      "step": 2029568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6764639210987164e-05,
+      "loss": 3.3977,
+      "step": 2030080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6756253263476644e-05,
+      "loss": 3.4084,
+      "step": 2030592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6747883694769857e-05,
+      "loss": 3.4002,
+      "step": 2031104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6739497747259337e-05,
+      "loss": 3.3998,
+      "step": 2031616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6731111799748817e-05,
+      "loss": 3.4161,
+      "step": 2032128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6722725852238293e-05,
+      "loss": 3.4164,
+      "step": 2032640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6714356283531506e-05,
+      "loss": 3.4082,
+      "step": 2033152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6705970336020986e-05,
+      "loss": 3.398,
+      "step": 2033664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.669758438851047e-05,
+      "loss": 3.4064,
+      "step": 2034176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6689214819803678e-05,
+      "loss": 3.3957,
+      "step": 2034688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6680828872293158e-05,
+      "loss": 3.4118,
+      "step": 2035200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6672442924782638e-05,
+      "loss": 3.4035,
+      "step": 2035712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6664056977272118e-05,
+      "loss": 3.407,
+      "step": 2036224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.665568740856533e-05,
+      "loss": 3.4023,
+      "step": 2036736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.664730146105481e-05,
+      "loss": 3.4056,
+      "step": 2037248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.663891551354429e-05,
+      "loss": 3.405,
+      "step": 2037760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6630529566033767e-05,
+      "loss": 3.4029,
+      "step": 2038272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.662215999732698e-05,
+      "loss": 3.3971,
+      "step": 2038784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.661377404981646e-05,
+      "loss": 3.402,
+      "step": 2039296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.660538810230594e-05,
+      "loss": 3.3939,
+      "step": 2039808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.659700215479542e-05,
+      "loss": 3.3962,
+      "step": 2040320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6588632586088632e-05,
+      "loss": 3.4043,
+      "step": 2040832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6580246638578112e-05,
+      "loss": 3.4003,
+      "step": 2041344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6571860691067592e-05,
+      "loss": 3.4048,
+      "step": 2041856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.656347474355707e-05,
+      "loss": 3.3959,
+      "step": 2042368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6555105174850284e-05,
+      "loss": 3.4148,
+      "step": 2042880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6546719227339764e-05,
+      "loss": 3.419,
+      "step": 2043392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.653833327982924e-05,
+      "loss": 3.4064,
+      "step": 2043904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.652994733231872e-05,
+      "loss": 3.4113,
+      "step": 2044416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6521577763611933e-05,
+      "loss": 3.3994,
+      "step": 2044928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6513208194905146e-05,
+      "loss": 3.3995,
+      "step": 2045440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6504822247394626e-05,
+      "loss": 3.4004,
+      "step": 2045952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6496436299884106e-05,
+      "loss": 3.3941,
+      "step": 2046464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6488050352373586e-05,
+      "loss": 3.4015,
+      "step": 2046976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6479664404863066e-05,
+      "loss": 3.4044,
+      "step": 2047488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6471278457352542e-05,
+      "loss": 3.4063,
+      "step": 2048000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6462892509842022e-05,
+      "loss": 3.4031,
+      "step": 2048512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6454506562331502e-05,
+      "loss": 3.4076,
+      "step": 2049024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6446136993624715e-05,
+      "loss": 3.4015,
+      "step": 2049536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6437751046114195e-05,
+      "loss": 3.4015,
+      "step": 2050048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6429365098603674e-05,
+      "loss": 3.4017,
+      "step": 2050560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6420979151093154e-05,
+      "loss": 3.4154,
+      "step": 2051072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6412609582386364e-05,
+      "loss": 3.4148,
+      "step": 2051584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6404223634875847e-05,
+      "loss": 3.3978,
+      "step": 2052096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6395837687365327e-05,
+      "loss": 3.4179,
+      "step": 2052608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6387451739854807e-05,
+      "loss": 3.4036,
+      "step": 2053120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6379082171148016e-05,
+      "loss": 3.4081,
+      "step": 2053632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6370696223637496e-05,
+      "loss": 3.407,
+      "step": 2054144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6362310276126976e-05,
+      "loss": 3.397,
+      "step": 2054656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6353924328616456e-05,
+      "loss": 3.4008,
+      "step": 2055168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6345554759909668e-05,
+      "loss": 3.4189,
+      "step": 2055680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6337168812399148e-05,
+      "loss": 3.3949,
+      "step": 2056192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6328782864888628e-05,
+      "loss": 3.4074,
+      "step": 2056704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6320413296181837e-05,
+      "loss": 3.4084,
+      "step": 2057216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6312027348671317e-05,
+      "loss": 3.4082,
+      "step": 2057728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.63036414011608e-05,
+      "loss": 3.3997,
+      "step": 2058240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.629525545365028e-05,
+      "loss": 3.4038,
+      "step": 2058752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.628688588494349e-05,
+      "loss": 3.4081,
+      "step": 2059264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.627849993743297e-05,
+      "loss": 3.4133,
+      "step": 2059776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.627011398992245e-05,
+      "loss": 3.405,
+      "step": 2060288
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8551113605499268,
+      "eval_runtime": 301.653,
+      "eval_samples_per_second": 1265.0,
+      "eval_steps_per_second": 39.532,
+      "step": 2060640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.626172804241193e-05,
+      "loss": 3.3787,
+      "step": 2060800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.625334209490141e-05,
+      "loss": 3.3996,
+      "step": 2061312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6244956147390893e-05,
+      "loss": 3.4074,
+      "step": 2061824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6236586578684102e-05,
+      "loss": 3.4088,
+      "step": 2062336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6228200631173582e-05,
+      "loss": 3.4181,
+      "step": 2062848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6219814683663062e-05,
+      "loss": 3.3994,
+      "step": 2063360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6211428736152542e-05,
+      "loss": 3.4051,
+      "step": 2063872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6203042788642018e-05,
+      "loss": 3.401,
+      "step": 2064384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.61946568411315e-05,
+      "loss": 3.3959,
+      "step": 2064896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.618627089362098e-05,
+      "loss": 3.3931,
+      "step": 2065408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.617790132491419e-05,
+      "loss": 3.4071,
+      "step": 2065920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.616951537740367e-05,
+      "loss": 3.4146,
+      "step": 2066432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.616112942989315e-05,
+      "loss": 3.4033,
+      "step": 2066944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.615274348238263e-05,
+      "loss": 3.3848,
+      "step": 2067456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6144373913675843e-05,
+      "loss": 3.4031,
+      "step": 2067968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6135987966165323e-05,
+      "loss": 3.3871,
+      "step": 2068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6127602018654803e-05,
+      "loss": 3.398,
+      "step": 2068992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6119216071144283e-05,
+      "loss": 3.4012,
+      "step": 2069504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6110846502437492e-05,
+      "loss": 3.4009,
+      "step": 2070016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6102460554926972e-05,
+      "loss": 3.404,
+      "step": 2070528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6094090986220188e-05,
+      "loss": 3.4123,
+      "step": 2071040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6085705038709664e-05,
+      "loss": 3.4059,
+      "step": 2071552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6077319091199144e-05,
+      "loss": 3.4032,
+      "step": 2072064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6068933143688624e-05,
+      "loss": 3.3944,
+      "step": 2072576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6060547196178104e-05,
+      "loss": 3.4017,
+      "step": 2073088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6052161248667584e-05,
+      "loss": 3.4028,
+      "step": 2073600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6043791679960797e-05,
+      "loss": 3.3969,
+      "step": 2074112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6035405732450277e-05,
+      "loss": 3.4006,
+      "step": 2074624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6027019784939757e-05,
+      "loss": 3.3932,
+      "step": 2075136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6018633837429237e-05,
+      "loss": 3.3997,
+      "step": 2075648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6010247889918717e-05,
+      "loss": 3.3933,
+      "step": 2076160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6001861942408196e-05,
+      "loss": 3.4067,
+      "step": 2076672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5993475994897676e-05,
+      "loss": 3.4028,
+      "step": 2077184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5985090047387156e-05,
+      "loss": 3.4057,
+      "step": 2077696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.597672047868037e-05,
+      "loss": 3.4005,
+      "step": 2078208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5968334531169845e-05,
+      "loss": 3.3987,
+      "step": 2078720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5959948583659325e-05,
+      "loss": 3.4071,
+      "step": 2079232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5951562636148805e-05,
+      "loss": 3.3939,
+      "step": 2079744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5943193067442018e-05,
+      "loss": 3.3912,
+      "step": 2080256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5934807119931498e-05,
+      "loss": 3.3936,
+      "step": 2080768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5926421172420978e-05,
+      "loss": 3.3921,
+      "step": 2081280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.591805160371419e-05,
+      "loss": 3.4008,
+      "step": 2081792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.590966565620367e-05,
+      "loss": 3.3945,
+      "step": 2082304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5901279708693147e-05,
+      "loss": 3.4013,
+      "step": 2082816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.589289376118263e-05,
+      "loss": 3.4,
+      "step": 2083328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5884524192475843e-05,
+      "loss": 3.406,
+      "step": 2083840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.587613824496532e-05,
+      "loss": 3.4052,
+      "step": 2084352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.58677522974548e-05,
+      "loss": 3.4003,
+      "step": 2084864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.585936634994428e-05,
+      "loss": 3.3738,
+      "step": 2085376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.585099678123749e-05,
+      "loss": 3.405,
+      "step": 2085888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.584261083372697e-05,
+      "loss": 3.3995,
+      "step": 2086400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.583422488621645e-05,
+      "loss": 3.3932,
+      "step": 2086912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.582583893870593e-05,
+      "loss": 3.398,
+      "step": 2087424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5817485748802873e-05,
+      "loss": 3.3955,
+      "step": 2087936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5809099801292353e-05,
+      "loss": 3.381,
+      "step": 2088448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5800713853781833e-05,
+      "loss": 3.3886,
+      "step": 2088960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5792327906271316e-05,
+      "loss": 3.394,
+      "step": 2089472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5783941958760793e-05,
+      "loss": 3.387,
+      "step": 2089984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5775556011250273e-05,
+      "loss": 3.4075,
+      "step": 2090496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5767170063739753e-05,
+      "loss": 3.3936,
+      "step": 2091008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5758784116229233e-05,
+      "loss": 3.384,
+      "step": 2091520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5750414547522442e-05,
+      "loss": 3.4059,
+      "step": 2092032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5742028600011925e-05,
+      "loss": 3.3897,
+      "step": 2092544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5733642652501405e-05,
+      "loss": 3.3806,
+      "step": 2093056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5725256704990885e-05,
+      "loss": 3.3986,
+      "step": 2093568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5716887136284094e-05,
+      "loss": 3.3925,
+      "step": 2094080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5708517567577307e-05,
+      "loss": 3.387,
+      "step": 2094592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5700131620066787e-05,
+      "loss": 3.396,
+      "step": 2095104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5691745672556267e-05,
+      "loss": 3.3808,
+      "step": 2095616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5683359725045747e-05,
+      "loss": 3.3909,
+      "step": 2096128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5674973777535226e-05,
+      "loss": 3.4028,
+      "step": 2096640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5666587830024706e-05,
+      "loss": 3.3948,
+      "step": 2097152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5658218261317916e-05,
+      "loss": 3.4051,
+      "step": 2097664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5649832313807395e-05,
+      "loss": 3.394,
+      "step": 2098176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.564144636629688e-05,
+      "loss": 3.4114,
+      "step": 2098688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.563306041878636e-05,
+      "loss": 3.3862,
+      "step": 2099200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.562467447127584e-05,
+      "loss": 3.403,
+      "step": 2099712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.561628852376532e-05,
+      "loss": 3.3975,
+      "step": 2100224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5607902576254795e-05,
+      "loss": 3.3871,
+      "step": 2100736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5599533007548008e-05,
+      "loss": 3.4034,
+      "step": 2101248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5591147060037488e-05,
+      "loss": 3.4013,
+      "step": 2101760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.558276111252697e-05,
+      "loss": 3.399,
+      "step": 2102272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5574375165016448e-05,
+      "loss": 3.4014,
+      "step": 2102784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.556600559630966e-05,
+      "loss": 3.3805,
+      "step": 2103296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.555761964879914e-05,
+      "loss": 3.3784,
+      "step": 2103808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.554923370128862e-05,
+      "loss": 3.3864,
+      "step": 2104320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.55408477537781e-05,
+      "loss": 3.3996,
+      "step": 2104832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5532478185071312e-05,
+      "loss": 3.391,
+      "step": 2105344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5524092237560792e-05,
+      "loss": 3.3927,
+      "step": 2105856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.551570629005027e-05,
+      "loss": 3.3889,
+      "step": 2106368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.550732034253975e-05,
+      "loss": 3.3935,
+      "step": 2106880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.549895077383296e-05,
+      "loss": 3.3921,
+      "step": 2107392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.549056482632244e-05,
+      "loss": 3.392,
+      "step": 2107904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.548217887881192e-05,
+      "loss": 3.4022,
+      "step": 2108416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.54737929313014e-05,
+      "loss": 3.4052,
+      "step": 2108928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5465423362594614e-05,
+      "loss": 3.4006,
+      "step": 2109440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5457037415084094e-05,
+      "loss": 3.3901,
+      "step": 2109952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.544865146757357e-05,
+      "loss": 3.3905,
+      "step": 2110464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.544026552006305e-05,
+      "loss": 3.3882,
+      "step": 2110976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5431895951356266e-05,
+      "loss": 3.4013,
+      "step": 2111488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5423510003845743e-05,
+      "loss": 3.3974,
+      "step": 2112000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5415124056335223e-05,
+      "loss": 3.3972,
+      "step": 2112512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5406738108824703e-05,
+      "loss": 3.3936,
+      "step": 2113024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5398368540117915e-05,
+      "loss": 3.3932,
+      "step": 2113536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5389982592607395e-05,
+      "loss": 3.3986,
+      "step": 2114048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5381596645096875e-05,
+      "loss": 3.3898,
+      "step": 2114560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5373210697586355e-05,
+      "loss": 3.3852,
+      "step": 2115072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5364841128879567e-05,
+      "loss": 3.3897,
+      "step": 2115584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5356455181369044e-05,
+      "loss": 3.3909,
+      "step": 2116096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5348069233858524e-05,
+      "loss": 3.3872,
+      "step": 2116608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5339683286348004e-05,
+      "loss": 3.3916,
+      "step": 2117120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5331313717641216e-05,
+      "loss": 3.392,
+      "step": 2117632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5322927770130696e-05,
+      "loss": 3.3951,
+      "step": 2118144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5314541822620176e-05,
+      "loss": 3.384,
+      "step": 2118656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5306155875109656e-05,
+      "loss": 3.4046,
+      "step": 2119168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5297786306402865e-05,
+      "loss": 3.41,
+      "step": 2119680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.528940035889235e-05,
+      "loss": 3.3958,
+      "step": 2120192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.528101441138183e-05,
+      "loss": 3.3991,
+      "step": 2120704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.527262846387131e-05,
+      "loss": 3.3916,
+      "step": 2121216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.526427527396825e-05,
+      "loss": 3.3905,
+      "step": 2121728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5255889326457732e-05,
+      "loss": 3.3927,
+      "step": 2122240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5247503378947212e-05,
+      "loss": 3.3826,
+      "step": 2122752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5239117431436688e-05,
+      "loss": 3.3938,
+      "step": 2123264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.523073148392617e-05,
+      "loss": 3.3945,
+      "step": 2123776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.522234553641565e-05,
+      "loss": 3.4026,
+      "step": 2124288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.521395958890513e-05,
+      "loss": 3.3878,
+      "step": 2124800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.520557364139461e-05,
+      "loss": 3.3996,
+      "step": 2125312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.519720407268782e-05,
+      "loss": 3.3907,
+      "step": 2125824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.51888181251773e-05,
+      "loss": 3.39,
+      "step": 2126336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.518043217766678e-05,
+      "loss": 3.3953,
+      "step": 2126848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5172046230156262e-05,
+      "loss": 3.3997,
+      "step": 2127360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5163676661449471e-05,
+      "loss": 3.4068,
+      "step": 2127872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5155290713938951e-05,
+      "loss": 3.3868,
+      "step": 2128384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5146904766428433e-05,
+      "loss": 3.4057,
+      "step": 2128896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5138518818917913e-05,
+      "loss": 3.3914,
+      "step": 2129408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5130149250211124e-05,
+      "loss": 3.399,
+      "step": 2129920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5121763302700604e-05,
+      "loss": 3.3981,
+      "step": 2130432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5113377355190084e-05,
+      "loss": 3.3878,
+      "step": 2130944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5104991407679564e-05,
+      "loss": 3.3917,
+      "step": 2131456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5096621838972774e-05,
+      "loss": 3.4092,
+      "step": 2131968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5088235891462254e-05,
+      "loss": 3.3829,
+      "step": 2132480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5079849943951734e-05,
+      "loss": 3.3971,
+      "step": 2132992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5071480375244945e-05,
+      "loss": 3.3958,
+      "step": 2133504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5063094427734425e-05,
+      "loss": 3.3954,
+      "step": 2134016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5054708480223905e-05,
+      "loss": 3.3927,
+      "step": 2134528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5046322532713387e-05,
+      "loss": 3.3918,
+      "step": 2135040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5037952964006596e-05,
+      "loss": 3.3998,
+      "step": 2135552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5029567016496077e-05,
+      "loss": 3.4037,
+      "step": 2136064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5021181068985557e-05,
+      "loss": 3.3944,
+      "step": 2136576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.857171058654785,
+      "eval_runtime": 307.9161,
+      "eval_samples_per_second": 1239.27,
+      "eval_steps_per_second": 38.728,
+      "step": 2136960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5012795121475037e-05,
+      "loss": 3.3878,
+      "step": 2137088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5004425552768248e-05,
+      "loss": 3.3851,
+      "step": 2137600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4996039605257728e-05,
+      "loss": 3.3962,
+      "step": 2138112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4987670036550939e-05,
+      "loss": 3.3999,
+      "step": 2138624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4979284089040419e-05,
+      "loss": 3.4027,
+      "step": 2139136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4970898141529899e-05,
+      "loss": 3.3919,
+      "step": 2139648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4962512194019379e-05,
+      "loss": 3.399,
+      "step": 2140160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4954126246508859e-05,
+      "loss": 3.388,
+      "step": 2140672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.494574029899834e-05,
+      "loss": 3.3857,
+      "step": 2141184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4937354351487817e-05,
+      "loss": 3.38,
+      "step": 2141696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4928968403977297e-05,
+      "loss": 3.4023,
+      "step": 2142208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4920598835270511e-05,
+      "loss": 3.4026,
+      "step": 2142720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4912212887759991e-05,
+      "loss": 3.3958,
+      "step": 2143232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.490382694024947e-05,
+      "loss": 3.3779,
+      "step": 2143744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4895457371542682e-05,
+      "loss": 3.3953,
+      "step": 2144256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4887071424032162e-05,
+      "loss": 3.3754,
+      "step": 2144768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.487868547652164e-05,
+      "loss": 3.3896,
+      "step": 2145280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.487029952901112e-05,
+      "loss": 3.3882,
+      "step": 2145792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4861929960304332e-05,
+      "loss": 3.389,
+      "step": 2146304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4853544012793812e-05,
+      "loss": 3.3942,
+      "step": 2146816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.484515806528329e-05,
+      "loss": 3.401,
+      "step": 2147328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.483677211777277e-05,
+      "loss": 3.3929,
+      "step": 2147840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4828402549065985e-05,
+      "loss": 3.3937,
+      "step": 2148352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4820016601555465e-05,
+      "loss": 3.3864,
+      "step": 2148864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4811630654044941e-05,
+      "loss": 3.3923,
+      "step": 2149376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4803244706534423e-05,
+      "loss": 3.3947,
+      "step": 2149888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4794875137827635e-05,
+      "loss": 3.3873,
+      "step": 2150400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4786489190317112e-05,
+      "loss": 3.3896,
+      "step": 2150912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4778103242806594e-05,
+      "loss": 3.3845,
+      "step": 2151424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4769717295296074e-05,
+      "loss": 3.3889,
+      "step": 2151936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4761347726589286e-05,
+      "loss": 3.3897,
+      "step": 2152448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4752961779078764e-05,
+      "loss": 3.3936,
+      "step": 2152960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4744575831568244e-05,
+      "loss": 3.388,
+      "step": 2153472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4736189884057724e-05,
+      "loss": 3.3999,
+      "step": 2153984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4727820315350937e-05,
+      "loss": 3.3909,
+      "step": 2154496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4719434367840415e-05,
+      "loss": 3.3863,
+      "step": 2155008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4711048420329895e-05,
+      "loss": 3.4,
+      "step": 2155520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4702662472819375e-05,
+      "loss": 3.3804,
+      "step": 2156032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4694292904112586e-05,
+      "loss": 3.3831,
+      "step": 2156544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4685906956602066e-05,
+      "loss": 3.3845,
+      "step": 2157056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4677521009091547e-05,
+      "loss": 3.3828,
+      "step": 2157568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4669135061581027e-05,
+      "loss": 3.3863,
+      "step": 2158080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4660765492874238e-05,
+      "loss": 3.3883,
+      "step": 2158592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4652379545363718e-05,
+      "loss": 3.3898,
+      "step": 2159104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4643993597853198e-05,
+      "loss": 3.39,
+      "step": 2159616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4635607650342678e-05,
+      "loss": 3.3939,
+      "step": 2160128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4627238081635889e-05,
+      "loss": 3.3955,
+      "step": 2160640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4618852134125369e-05,
+      "loss": 3.3909,
+      "step": 2161152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4610466186614849e-05,
+      "loss": 3.368,
+      "step": 2161664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4602080239104329e-05,
+      "loss": 3.3947,
+      "step": 2162176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.459371067039754e-05,
+      "loss": 3.3861,
+      "step": 2162688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.458532472288702e-05,
+      "loss": 3.3871,
+      "step": 2163200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4576938775376501e-05,
+      "loss": 3.3828,
+      "step": 2163712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.456856920666971e-05,
+      "loss": 3.383,
+      "step": 2164224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.456018325915919e-05,
+      "loss": 3.3745,
+      "step": 2164736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4551797311648672e-05,
+      "loss": 3.379,
+      "step": 2165248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4543411364138152e-05,
+      "loss": 3.3857,
+      "step": 2165760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4535025416627632e-05,
+      "loss": 3.3754,
+      "step": 2166272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4526639469117112e-05,
+      "loss": 3.3989,
+      "step": 2166784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4518253521606593e-05,
+      "loss": 3.3847,
+      "step": 2167296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.450986757409607e-05,
+      "loss": 3.3694,
+      "step": 2167808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4501498005389282e-05,
+      "loss": 3.3992,
+      "step": 2168320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4493112057878764e-05,
+      "loss": 3.3768,
+      "step": 2168832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.448472611036824e-05,
+      "loss": 3.3719,
+      "step": 2169344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.447634016285772e-05,
+      "loss": 3.3867,
+      "step": 2169856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4467970594150935e-05,
+      "loss": 3.3819,
+      "step": 2170368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4459584646640415e-05,
+      "loss": 3.3767,
+      "step": 2170880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4451198699129893e-05,
+      "loss": 3.3881,
+      "step": 2171392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4442812751619373e-05,
+      "loss": 3.3749,
+      "step": 2171904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4434443182912585e-05,
+      "loss": 3.3785,
+      "step": 2172416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4426057235402065e-05,
+      "loss": 3.3943,
+      "step": 2172928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4417671287891543e-05,
+      "loss": 3.3842,
+      "step": 2173440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4409301719184756e-05,
+      "loss": 3.3955,
+      "step": 2173952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4400915771674236e-05,
+      "loss": 3.3868,
+      "step": 2174464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4392529824163714e-05,
+      "loss": 3.4039,
+      "step": 2174976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4384143876653194e-05,
+      "loss": 3.377,
+      "step": 2175488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4375757929142674e-05,
+      "loss": 3.3952,
+      "step": 2176000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4367371981632156e-05,
+      "loss": 3.3832,
+      "step": 2176512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4358986034121636e-05,
+      "loss": 3.3809,
+      "step": 2177024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4350600086611116e-05,
+      "loss": 3.388,
+      "step": 2177536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4342230517904326e-05,
+      "loss": 3.397,
+      "step": 2178048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4333844570393806e-05,
+      "loss": 3.3865,
+      "step": 2178560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4325458622883286e-05,
+      "loss": 3.3935,
+      "step": 2179072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4317089054176497e-05,
+      "loss": 3.3725,
+      "step": 2179584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4308703106665977e-05,
+      "loss": 3.3645,
+      "step": 2180096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4300317159155457e-05,
+      "loss": 3.3772,
+      "step": 2180608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4291931211644937e-05,
+      "loss": 3.3861,
+      "step": 2181120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4283561642938148e-05,
+      "loss": 3.3863,
+      "step": 2181632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4275175695427628e-05,
+      "loss": 3.3827,
+      "step": 2182144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.426678974791711e-05,
+      "loss": 3.3785,
+      "step": 2182656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.425840380040659e-05,
+      "loss": 3.3862,
+      "step": 2183168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4250034231699798e-05,
+      "loss": 3.3815,
+      "step": 2183680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.424164828418928e-05,
+      "loss": 3.3807,
+      "step": 2184192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.423326233667876e-05,
+      "loss": 3.3928,
+      "step": 2184704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.422487638916824e-05,
+      "loss": 3.3938,
+      "step": 2185216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.421650682046145e-05,
+      "loss": 3.3945,
+      "step": 2185728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.420812087295093e-05,
+      "loss": 3.3761,
+      "step": 2186240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.419973492544041e-05,
+      "loss": 3.3856,
+      "step": 2186752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.419134897792989e-05,
+      "loss": 3.3777,
+      "step": 2187264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4182979409223101e-05,
+      "loss": 3.3901,
+      "step": 2187776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4174593461712581e-05,
+      "loss": 3.388,
+      "step": 2188288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4166207514202063e-05,
+      "loss": 3.3838,
+      "step": 2188800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4157821566691543e-05,
+      "loss": 3.384,
+      "step": 2189312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4149451997984752e-05,
+      "loss": 3.3854,
+      "step": 2189824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4141066050474234e-05,
+      "loss": 3.3873,
+      "step": 2190336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4132680102963714e-05,
+      "loss": 3.381,
+      "step": 2190848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.412429415545319e-05,
+      "loss": 3.3795,
+      "step": 2191360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4115924586746404e-05,
+      "loss": 3.3795,
+      "step": 2191872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4107538639235884e-05,
+      "loss": 3.3793,
+      "step": 2192384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4099152691725364e-05,
+      "loss": 3.3782,
+      "step": 2192896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4090766744214843e-05,
+      "loss": 3.3799,
+      "step": 2193408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4082397175508055e-05,
+      "loss": 3.3811,
+      "step": 2193920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4074011227997535e-05,
+      "loss": 3.3883,
+      "step": 2194432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4065625280487017e-05,
+      "loss": 3.3731,
+      "step": 2194944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4057239332976493e-05,
+      "loss": 3.3938,
+      "step": 2195456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4048869764269706e-05,
+      "loss": 3.3997,
+      "step": 2195968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4040483816759187e-05,
+      "loss": 3.3893,
+      "step": 2196480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4032097869248664e-05,
+      "loss": 3.3871,
+      "step": 2196992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4023711921738144e-05,
+      "loss": 3.3877,
+      "step": 2197504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4015358731835087e-05,
+      "loss": 3.3757,
+      "step": 2198016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4006972784324567e-05,
+      "loss": 3.3879,
+      "step": 2198528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3998586836814049e-05,
+      "loss": 3.3686,
+      "step": 2199040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3990200889303529e-05,
+      "loss": 3.3854,
+      "step": 2199552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3981814941793009e-05,
+      "loss": 3.3838,
+      "step": 2200064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3973428994282489e-05,
+      "loss": 3.3885,
+      "step": 2200576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3965043046771967e-05,
+      "loss": 3.3822,
+      "step": 2201088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3956657099261447e-05,
+      "loss": 3.3888,
+      "step": 2201600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.394828753055466e-05,
+      "loss": 3.3802,
+      "step": 2202112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3939901583044138e-05,
+      "loss": 3.3768,
+      "step": 2202624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3931515635533618e-05,
+      "loss": 3.3882,
+      "step": 2203136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3923129688023098e-05,
+      "loss": 3.3883,
+      "step": 2203648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3914760119316312e-05,
+      "loss": 3.398,
+      "step": 2204160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3906374171805788e-05,
+      "loss": 3.3812,
+      "step": 2204672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.389798822429527e-05,
+      "loss": 3.3937,
+      "step": 2205184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.388960227678475e-05,
+      "loss": 3.3821,
+      "step": 2205696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3881232708077962e-05,
+      "loss": 3.3861,
+      "step": 2206208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.387284676056744e-05,
+      "loss": 3.39,
+      "step": 2206720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.386446081305692e-05,
+      "loss": 3.382,
+      "step": 2207232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.38560748655464e-05,
+      "loss": 3.3823,
+      "step": 2207744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3847705296839611e-05,
+      "loss": 3.3968,
+      "step": 2208256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3839319349329091e-05,
+      "loss": 3.3764,
+      "step": 2208768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3830933401818571e-05,
+      "loss": 3.3858,
+      "step": 2209280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3822563833111784e-05,
+      "loss": 3.3895,
+      "step": 2209792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3814177885601262e-05,
+      "loss": 3.3862,
+      "step": 2210304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3805791938090742e-05,
+      "loss": 3.3837,
+      "step": 2210816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3797405990580222e-05,
+      "loss": 3.3848,
+      "step": 2211328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3789036421873436e-05,
+      "loss": 3.3862,
+      "step": 2211840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3780650474362913e-05,
+      "loss": 3.3933,
+      "step": 2212352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3772264526852394e-05,
+      "loss": 3.3867,
+      "step": 2212864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8572282791137695,
+      "eval_runtime": 308.2173,
+      "eval_samples_per_second": 1238.058,
+      "eval_steps_per_second": 38.69,
+      "step": 2213280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3763878579341874e-05,
+      "loss": 3.3702,
+      "step": 2213376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3755492631831354e-05,
+      "loss": 3.3745,
+      "step": 2213888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3747106684320834e-05,
+      "loss": 3.3845,
+      "step": 2214400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3738737115614045e-05,
+      "loss": 3.3957,
+      "step": 2214912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3730351168103525e-05,
+      "loss": 3.3886,
+      "step": 2215424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3721965220593005e-05,
+      "loss": 3.3833,
+      "step": 2215936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3713579273082487e-05,
+      "loss": 3.3848,
+      "step": 2216448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3705193325571967e-05,
+      "loss": 3.3802,
+      "step": 2216960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3696807378061443e-05,
+      "loss": 3.3728,
+      "step": 2217472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3688421430550925e-05,
+      "loss": 3.3759,
+      "step": 2217984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3680035483040405e-05,
+      "loss": 3.3892,
+      "step": 2218496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3671665914333614e-05,
+      "loss": 3.3908,
+      "step": 2219008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3663279966823095e-05,
+      "loss": 3.3917,
+      "step": 2219520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3654894019312575e-05,
+      "loss": 3.3614,
+      "step": 2220032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3646508071802055e-05,
+      "loss": 3.3782,
+      "step": 2220544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3638138503095266e-05,
+      "loss": 3.3688,
+      "step": 2221056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3629752555584746e-05,
+      "loss": 3.3808,
+      "step": 2221568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3621366608074226e-05,
+      "loss": 3.3775,
+      "step": 2222080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.361299703936744e-05,
+      "loss": 3.3825,
+      "step": 2222592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3604611091856917e-05,
+      "loss": 3.3787,
+      "step": 2223104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.359624152315013e-05,
+      "loss": 3.3953,
+      "step": 2223616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3587855575639611e-05,
+      "loss": 3.3802,
+      "step": 2224128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3579469628129088e-05,
+      "loss": 3.3848,
+      "step": 2224640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3571083680618567e-05,
+      "loss": 3.3759,
+      "step": 2225152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3562697733108049e-05,
+      "loss": 3.3842,
+      "step": 2225664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3554311785597529e-05,
+      "loss": 3.3841,
+      "step": 2226176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.354594221689074e-05,
+      "loss": 3.3756,
+      "step": 2226688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.353755626938022e-05,
+      "loss": 3.381,
+      "step": 2227200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.35291703218697e-05,
+      "loss": 3.3733,
+      "step": 2227712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.352078437435918e-05,
+      "loss": 3.3756,
+      "step": 2228224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.351239842684866e-05,
+      "loss": 3.3833,
+      "step": 2228736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3504012479338141e-05,
+      "loss": 3.3818,
+      "step": 2229248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3495626531827621e-05,
+      "loss": 3.3821,
+      "step": 2229760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3487240584317098e-05,
+      "loss": 3.387,
+      "step": 2230272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3478871015610312e-05,
+      "loss": 3.3833,
+      "step": 2230784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3470485068099792e-05,
+      "loss": 3.3779,
+      "step": 2231296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3462099120589269e-05,
+      "loss": 3.3916,
+      "step": 2231808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.345371317307875e-05,
+      "loss": 3.3666,
+      "step": 2232320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3445343604371963e-05,
+      "loss": 3.376,
+      "step": 2232832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3436957656861443e-05,
+      "loss": 3.3762,
+      "step": 2233344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3428571709350921e-05,
+      "loss": 3.3682,
+      "step": 2233856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3420185761840401e-05,
+      "loss": 3.3791,
+      "step": 2234368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3411816193133613e-05,
+      "loss": 3.3821,
+      "step": 2234880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3403430245623095e-05,
+      "loss": 3.38,
+      "step": 2235392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3395044298112572e-05,
+      "loss": 3.3802,
+      "step": 2235904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3386658350602052e-05,
+      "loss": 3.3832,
+      "step": 2236416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3378288781895266e-05,
+      "loss": 3.3871,
+      "step": 2236928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3369902834384742e-05,
+      "loss": 3.3796,
+      "step": 2237440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3361516886874222e-05,
+      "loss": 3.3621,
+      "step": 2237952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3353147318167436e-05,
+      "loss": 3.3804,
+      "step": 2238464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3344761370656916e-05,
+      "loss": 3.3801,
+      "step": 2238976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3336375423146395e-05,
+      "loss": 3.3819,
+      "step": 2239488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3327989475635875e-05,
+      "loss": 3.3731,
+      "step": 2240000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3319636285732818e-05,
+      "loss": 3.3686,
+      "step": 2240512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3311250338222298e-05,
+      "loss": 3.366,
+      "step": 2241024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3302864390711778e-05,
+      "loss": 3.3705,
+      "step": 2241536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3294478443201258e-05,
+      "loss": 3.3784,
+      "step": 2242048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3286092495690738e-05,
+      "loss": 3.3655,
+      "step": 2242560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3277706548180216e-05,
+      "loss": 3.388,
+      "step": 2243072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3269320600669696e-05,
+      "loss": 3.3777,
+      "step": 2243584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3260934653159176e-05,
+      "loss": 3.3586,
+      "step": 2244096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.325256508445239e-05,
+      "loss": 3.388,
+      "step": 2244608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3244179136941867e-05,
+      "loss": 3.3679,
+      "step": 2245120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3235793189431348e-05,
+      "loss": 3.3687,
+      "step": 2245632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3227407241920828e-05,
+      "loss": 3.3685,
+      "step": 2246144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.321903767321404e-05,
+      "loss": 3.377,
+      "step": 2246656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3210651725703519e-05,
+      "loss": 3.3655,
+      "step": 2247168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3202265778192999e-05,
+      "loss": 3.3786,
+      "step": 2247680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3193879830682479e-05,
+      "loss": 3.3633,
+      "step": 2248192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.318551026197569e-05,
+      "loss": 3.3688,
+      "step": 2248704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.317712431446517e-05,
+      "loss": 3.387,
+      "step": 2249216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.316873836695465e-05,
+      "loss": 3.3737,
+      "step": 2249728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3160368798247864e-05,
+      "loss": 3.3844,
+      "step": 2250240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.315198285073734e-05,
+      "loss": 3.376,
+      "step": 2250752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.314359690322682e-05,
+      "loss": 3.3929,
+      "step": 2251264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3135210955716302e-05,
+      "loss": 3.3684,
+      "step": 2251776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3126825008205782e-05,
+      "loss": 3.3876,
+      "step": 2252288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3118439060695262e-05,
+      "loss": 3.3697,
+      "step": 2252800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3110053113184742e-05,
+      "loss": 3.3701,
+      "step": 2253312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3101667165674222e-05,
+      "loss": 3.3781,
+      "step": 2253824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3093297596967433e-05,
+      "loss": 3.3873,
+      "step": 2254336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3084911649456913e-05,
+      "loss": 3.3783,
+      "step": 2254848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3076525701946392e-05,
+      "loss": 3.3827,
+      "step": 2255360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.306813975443587e-05,
+      "loss": 3.3634,
+      "step": 2255872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3059770185729083e-05,
+      "loss": 3.3602,
+      "step": 2256384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3051384238218565e-05,
+      "loss": 3.3703,
+      "step": 2256896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3042998290708045e-05,
+      "loss": 3.3737,
+      "step": 2257408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3034628722001254e-05,
+      "loss": 3.3726,
+      "step": 2257920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3026242774490736e-05,
+      "loss": 3.374,
+      "step": 2258432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3017856826980216e-05,
+      "loss": 3.3736,
+      "step": 2258944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3009470879469692e-05,
+      "loss": 3.3721,
+      "step": 2259456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3001101310762906e-05,
+      "loss": 3.3756,
+      "step": 2259968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2992715363252386e-05,
+      "loss": 3.3706,
+      "step": 2260480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2984329415741866e-05,
+      "loss": 3.3823,
+      "step": 2260992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2975943468231344e-05,
+      "loss": 3.3853,
+      "step": 2261504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2967573899524557e-05,
+      "loss": 3.3832,
+      "step": 2262016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2959187952014037e-05,
+      "loss": 3.3661,
+      "step": 2262528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2950802004503519e-05,
+      "loss": 3.3718,
+      "step": 2263040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2942416056992995e-05,
+      "loss": 3.3695,
+      "step": 2263552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2934046488286208e-05,
+      "loss": 3.3824,
+      "step": 2264064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.292566054077569e-05,
+      "loss": 3.3739,
+      "step": 2264576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2917274593265166e-05,
+      "loss": 3.375,
+      "step": 2265088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2908888645754646e-05,
+      "loss": 3.3734,
+      "step": 2265600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.290051907704786e-05,
+      "loss": 3.3806,
+      "step": 2266112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.289213312953734e-05,
+      "loss": 3.3767,
+      "step": 2266624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2883747182026818e-05,
+      "loss": 3.3734,
+      "step": 2267136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2875361234516298e-05,
+      "loss": 3.3697,
+      "step": 2267648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.286699166580951e-05,
+      "loss": 3.3661,
+      "step": 2268160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.285860571829899e-05,
+      "loss": 3.3722,
+      "step": 2268672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2850219770788469e-05,
+      "loss": 3.366,
+      "step": 2269184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2841833823277949e-05,
+      "loss": 3.3707,
+      "step": 2269696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2833464254571161e-05,
+      "loss": 3.3691,
+      "step": 2270208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.282507830706064e-05,
+      "loss": 3.3773,
+      "step": 2270720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.281669235955012e-05,
+      "loss": 3.3701,
+      "step": 2271232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.28083064120396e-05,
+      "loss": 3.3784,
+      "step": 2271744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2799936843332814e-05,
+      "loss": 3.3873,
+      "step": 2272256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.279155089582229e-05,
+      "loss": 3.3829,
+      "step": 2272768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2783164948311772e-05,
+      "loss": 3.3819,
+      "step": 2273280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2774779000801252e-05,
+      "loss": 3.3743,
+      "step": 2273792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2766425810898195e-05,
+      "loss": 3.374,
+      "step": 2274304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2758039863387675e-05,
+      "loss": 3.3726,
+      "step": 2274816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2749653915877155e-05,
+      "loss": 3.3612,
+      "step": 2275328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2741267968366635e-05,
+      "loss": 3.376,
+      "step": 2275840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2732882020856113e-05,
+      "loss": 3.3692,
+      "step": 2276352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2724496073345593e-05,
+      "loss": 3.3818,
+      "step": 2276864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2716110125835073e-05,
+      "loss": 3.3763,
+      "step": 2277376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2707724178324553e-05,
+      "loss": 3.376,
+      "step": 2277888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2699354609617764e-05,
+      "loss": 3.3686,
+      "step": 2278400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2690968662107244e-05,
+      "loss": 3.3713,
+      "step": 2278912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2682582714596724e-05,
+      "loss": 3.3777,
+      "step": 2279424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2674196767086205e-05,
+      "loss": 3.3785,
+      "step": 2279936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2665827198379415e-05,
+      "loss": 3.3898,
+      "step": 2280448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2657441250868896e-05,
+      "loss": 3.3708,
+      "step": 2280960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2649055303358376e-05,
+      "loss": 3.3815,
+      "step": 2281472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2640669355847856e-05,
+      "loss": 3.3744,
+      "step": 2281984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2632299787141067e-05,
+      "loss": 3.376,
+      "step": 2282496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2623913839630547e-05,
+      "loss": 3.3816,
+      "step": 2283008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2615527892120027e-05,
+      "loss": 3.3715,
+      "step": 2283520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2607141944609507e-05,
+      "loss": 3.3734,
+      "step": 2284032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2598772375902718e-05,
+      "loss": 3.3854,
+      "step": 2284544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2590386428392198e-05,
+      "loss": 3.3722,
+      "step": 2285056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2582000480881677e-05,
+      "loss": 3.371,
+      "step": 2285568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2573630912174888e-05,
+      "loss": 3.3793,
+      "step": 2286080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2565244964664368e-05,
+      "loss": 3.3807,
+      "step": 2286592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.255685901715385e-05,
+      "loss": 3.3736,
+      "step": 2287104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.254847306964333e-05,
+      "loss": 3.3759,
+      "step": 2287616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2540103500936539e-05,
+      "loss": 3.3724,
+      "step": 2288128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.253171755342602e-05,
+      "loss": 3.3867,
+      "step": 2288640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.25233316059155e-05,
+      "loss": 3.3758,
+      "step": 2289152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8584554195404053,
+      "eval_runtime": 307.1972,
+      "eval_samples_per_second": 1242.17,
+      "eval_steps_per_second": 38.819,
+      "step": 2289600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.251494565840498e-05,
+      "loss": 3.3692,
+      "step": 2289664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2506576089698191e-05,
+      "loss": 3.3672,
+      "step": 2290176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2498190142187671e-05,
+      "loss": 3.373,
+      "step": 2290688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2489804194677151e-05,
+      "loss": 3.3883,
+      "step": 2291200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2481418247166631e-05,
+      "loss": 3.3817,
+      "step": 2291712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2473048678459842e-05,
+      "loss": 3.3749,
+      "step": 2292224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2464662730949322e-05,
+      "loss": 3.3722,
+      "step": 2292736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2456276783438804e-05,
+      "loss": 3.3716,
+      "step": 2293248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2447890835928283e-05,
+      "loss": 3.3605,
+      "step": 2293760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2439521267221493e-05,
+      "loss": 3.3715,
+      "step": 2294272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2431135319710974e-05,
+      "loss": 3.3739,
+      "step": 2294784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2422749372200454e-05,
+      "loss": 3.3852,
+      "step": 2295296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2414363424689932e-05,
+      "loss": 3.3803,
+      "step": 2295808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2405993855983145e-05,
+      "loss": 3.3553,
+      "step": 2296320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2397607908472625e-05,
+      "loss": 3.3727,
+      "step": 2296832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2389221960962105e-05,
+      "loss": 3.3593,
+      "step": 2297344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2380836013451585e-05,
+      "loss": 3.3698,
+      "step": 2297856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2372466444744796e-05,
+      "loss": 3.3685,
+      "step": 2298368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2364080497234276e-05,
+      "loss": 3.3745,
+      "step": 2298880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2355694549723756e-05,
+      "loss": 3.367,
+      "step": 2299392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2347324981016966e-05,
+      "loss": 3.3864,
+      "step": 2299904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2338939033506446e-05,
+      "loss": 3.3719,
+      "step": 2300416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2330553085995928e-05,
+      "loss": 3.375,
+      "step": 2300928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2322167138485406e-05,
+      "loss": 3.3661,
+      "step": 2301440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2313781190974886e-05,
+      "loss": 3.3795,
+      "step": 2301952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2305395243464366e-05,
+      "loss": 3.3703,
+      "step": 2302464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2297025674757579e-05,
+      "loss": 3.3686,
+      "step": 2302976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2288639727247057e-05,
+      "loss": 3.3709,
+      "step": 2303488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2280253779736538e-05,
+      "loss": 3.3665,
+      "step": 2304000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2271867832226018e-05,
+      "loss": 3.3653,
+      "step": 2304512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2263481884715497e-05,
+      "loss": 3.3721,
+      "step": 2305024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2255095937204977e-05,
+      "loss": 3.3691,
+      "step": 2305536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2246709989694458e-05,
+      "loss": 3.3738,
+      "step": 2306048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2238324042183937e-05,
+      "loss": 3.3757,
+      "step": 2306560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2229954473477147e-05,
+      "loss": 3.3748,
+      "step": 2307072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2221568525966629e-05,
+      "loss": 3.3739,
+      "step": 2307584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2213182578456109e-05,
+      "loss": 3.3753,
+      "step": 2308096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2204796630945587e-05,
+      "loss": 3.3576,
+      "step": 2308608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.21964270622388e-05,
+      "loss": 3.3673,
+      "step": 2309120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.218804111472828e-05,
+      "loss": 3.3654,
+      "step": 2309632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.217965516721776e-05,
+      "loss": 3.3586,
+      "step": 2310144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.217126921970724e-05,
+      "loss": 3.3696,
+      "step": 2310656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.216289965100045e-05,
+      "loss": 3.376,
+      "step": 2311168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.215451370348993e-05,
+      "loss": 3.3656,
+      "step": 2311680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.214612775597941e-05,
+      "loss": 3.3731,
+      "step": 2312192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.213774180846889e-05,
+      "loss": 3.3726,
+      "step": 2312704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2129372239762101e-05,
+      "loss": 3.3781,
+      "step": 2313216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2120986292251583e-05,
+      "loss": 3.3702,
+      "step": 2313728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2112600344741061e-05,
+      "loss": 3.3518,
+      "step": 2314240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2104214397230541e-05,
+      "loss": 3.3712,
+      "step": 2314752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2095844828523753e-05,
+      "loss": 3.3727,
+      "step": 2315264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2087458881013233e-05,
+      "loss": 3.3717,
+      "step": 2315776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2079072933502712e-05,
+      "loss": 3.3642,
+      "step": 2316288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2070686985992193e-05,
+      "loss": 3.3603,
+      "step": 2316800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2062317417285404e-05,
+      "loss": 3.3593,
+      "step": 2317312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2053931469774884e-05,
+      "loss": 3.3589,
+      "step": 2317824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2045545522264364e-05,
+      "loss": 3.3684,
+      "step": 2318336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2037175953557575e-05,
+      "loss": 3.3538,
+      "step": 2318848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2028790006047055e-05,
+      "loss": 3.3791,
+      "step": 2319360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2020404058536535e-05,
+      "loss": 3.3689,
+      "step": 2319872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2012018111026015e-05,
+      "loss": 3.3522,
+      "step": 2320384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2003648542319227e-05,
+      "loss": 3.3783,
+      "step": 2320896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1995262594808707e-05,
+      "loss": 3.3573,
+      "step": 2321408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1986876647298185e-05,
+      "loss": 3.3626,
+      "step": 2321920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1978490699787665e-05,
+      "loss": 3.3589,
+      "step": 2322432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1970121131080878e-05,
+      "loss": 3.3661,
+      "step": 2322944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1961735183570356e-05,
+      "loss": 3.3578,
+      "step": 2323456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1953349236059836e-05,
+      "loss": 3.3696,
+      "step": 2323968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1944963288549318e-05,
+      "loss": 3.3507,
+      "step": 2324480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1936593719842528e-05,
+      "loss": 3.3622,
+      "step": 2324992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1928207772332008e-05,
+      "loss": 3.3756,
+      "step": 2325504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1919821824821488e-05,
+      "loss": 3.3669,
+      "step": 2326016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1911435877310968e-05,
+      "loss": 3.372,
+      "step": 2326528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.190306630860418e-05,
+      "loss": 3.368,
+      "step": 2327040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1894680361093659e-05,
+      "loss": 3.3825,
+      "step": 2327552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1886294413583139e-05,
+      "loss": 3.3607,
+      "step": 2328064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1877908466072619e-05,
+      "loss": 3.3797,
+      "step": 2328576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.186953889736583e-05,
+      "loss": 3.3585,
+      "step": 2329088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.186115294985531e-05,
+      "loss": 3.3636,
+      "step": 2329600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.185276700234479e-05,
+      "loss": 3.3695,
+      "step": 2330112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1844381054834271e-05,
+      "loss": 3.3746,
+      "step": 2330624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.183601148612748e-05,
+      "loss": 3.3707,
+      "step": 2331136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1827625538616962e-05,
+      "loss": 3.3719,
+      "step": 2331648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1819239591106442e-05,
+      "loss": 3.3548,
+      "step": 2332160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.181085364359592e-05,
+      "loss": 3.348,
+      "step": 2332672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1802484074889133e-05,
+      "loss": 3.3617,
+      "step": 2333184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1794098127378613e-05,
+      "loss": 3.3641,
+      "step": 2333696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1785712179868093e-05,
+      "loss": 3.3639,
+      "step": 2334208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1777326232357571e-05,
+      "loss": 3.3607,
+      "step": 2334720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1768956663650783e-05,
+      "loss": 3.3659,
+      "step": 2335232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1760570716140263e-05,
+      "loss": 3.3607,
+      "step": 2335744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1752184768629743e-05,
+      "loss": 3.3644,
+      "step": 2336256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1743798821119223e-05,
+      "loss": 3.3607,
+      "step": 2336768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1735429252412434e-05,
+      "loss": 3.3712,
+      "step": 2337280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1727043304901916e-05,
+      "loss": 3.3753,
+      "step": 2337792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1718657357391394e-05,
+      "loss": 3.3752,
+      "step": 2338304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1710271409880874e-05,
+      "loss": 3.3553,
+      "step": 2338816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1701901841174086e-05,
+      "loss": 3.3644,
+      "step": 2339328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1693515893663566e-05,
+      "loss": 3.3606,
+      "step": 2339840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1685129946153045e-05,
+      "loss": 3.3692,
+      "step": 2340352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1676743998642525e-05,
+      "loss": 3.3694,
+      "step": 2340864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1668374429935737e-05,
+      "loss": 3.3655,
+      "step": 2341376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1659988482425215e-05,
+      "loss": 3.3642,
+      "step": 2341888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1651602534914697e-05,
+      "loss": 3.3657,
+      "step": 2342400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1643216587404177e-05,
+      "loss": 3.3686,
+      "step": 2342912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1634847018697388e-05,
+      "loss": 3.3607,
+      "step": 2343424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1626461071186868e-05,
+      "loss": 3.3656,
+      "step": 2343936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1618075123676348e-05,
+      "loss": 3.3556,
+      "step": 2344448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1609705554969558e-05,
+      "loss": 3.3641,
+      "step": 2344960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.160131960745904e-05,
+      "loss": 3.3544,
+      "step": 2345472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1592933659948518e-05,
+      "loss": 3.3592,
+      "step": 2345984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1584547712437998e-05,
+      "loss": 3.3596,
+      "step": 2346496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.157617814373121e-05,
+      "loss": 3.3675,
+      "step": 2347008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1567792196220689e-05,
+      "loss": 3.3597,
+      "step": 2347520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1559406248710169e-05,
+      "loss": 3.3681,
+      "step": 2348032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.155102030119965e-05,
+      "loss": 3.379,
+      "step": 2348544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1542650732492861e-05,
+      "loss": 3.3763,
+      "step": 2349056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.153426478498234e-05,
+      "loss": 3.3719,
+      "step": 2349568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1525878837471821e-05,
+      "loss": 3.364,
+      "step": 2350080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1517509268765032e-05,
+      "loss": 3.365,
+      "step": 2350592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1509123321254512e-05,
+      "loss": 3.3606,
+      "step": 2351104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1500737373743992e-05,
+      "loss": 3.3528,
+      "step": 2351616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1492351426233472e-05,
+      "loss": 3.3673,
+      "step": 2352128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1483965478722952e-05,
+      "loss": 3.3603,
+      "step": 2352640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1475579531212432e-05,
+      "loss": 3.3731,
+      "step": 2353152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1467193583701912e-05,
+      "loss": 3.3604,
+      "step": 2353664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1458807636191392e-05,
+      "loss": 3.3666,
+      "step": 2354176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1450438067484604e-05,
+      "loss": 3.3608,
+      "step": 2354688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1442052119974083e-05,
+      "loss": 3.3604,
+      "step": 2355200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1433666172463563e-05,
+      "loss": 3.3689,
+      "step": 2355712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1425280224953042e-05,
+      "loss": 3.3669,
+      "step": 2356224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1416910656246253e-05,
+      "loss": 3.3824,
+      "step": 2356736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1408524708735733e-05,
+      "loss": 3.3635,
+      "step": 2357248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1400138761225213e-05,
+      "loss": 3.3693,
+      "step": 2357760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1391752813714695e-05,
+      "loss": 3.3676,
+      "step": 2358272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1383383245007904e-05,
+      "loss": 3.367,
+      "step": 2358784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1374997297497386e-05,
+      "loss": 3.3681,
+      "step": 2359296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1366611349986866e-05,
+      "loss": 3.3597,
+      "step": 2359808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1358225402476344e-05,
+      "loss": 3.3707,
+      "step": 2360320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1349855833769556e-05,
+      "loss": 3.3714,
+      "step": 2360832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1341469886259036e-05,
+      "loss": 3.3637,
+      "step": 2361344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1333083938748516e-05,
+      "loss": 3.3626,
+      "step": 2361856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1324697991237994e-05,
+      "loss": 3.3668,
+      "step": 2362368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1316328422531207e-05,
+      "loss": 3.3693,
+      "step": 2362880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1307942475020687e-05,
+      "loss": 3.3625,
+      "step": 2363392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1299556527510167e-05,
+      "loss": 3.3685,
+      "step": 2363904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1291170579999647e-05,
+      "loss": 3.3655,
+      "step": 2364416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1282801011292858e-05,
+      "loss": 3.3816,
+      "step": 2364928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.127441506378234e-05,
+      "loss": 3.3616,
+      "step": 2365440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.859992742538452,
+      "eval_runtime": 306.4447,
+      "eval_samples_per_second": 1245.22,
+      "eval_steps_per_second": 38.914,
+      "step": 2365920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1266029116271817e-05,
+      "loss": 3.3271,
+      "step": 2365952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1257643168761297e-05,
+      "loss": 3.3591,
+      "step": 2366464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1249257221250777e-05,
+      "loss": 3.3613,
+      "step": 2366976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1240871273740259e-05,
+      "loss": 3.3783,
+      "step": 2367488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1232501705033468e-05,
+      "loss": 3.3708,
+      "step": 2368000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1224115757522948e-05,
+      "loss": 3.3694,
+      "step": 2368512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.121572981001243e-05,
+      "loss": 3.3607,
+      "step": 2369024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1207343862501908e-05,
+      "loss": 3.3636,
+      "step": 2369536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.119897429379512e-05,
+      "loss": 3.3538,
+      "step": 2370048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.11905883462846e-05,
+      "loss": 3.3625,
+      "step": 2370560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.118220239877408e-05,
+      "loss": 3.3652,
+      "step": 2371072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1173816451263559e-05,
+      "loss": 3.3761,
+      "step": 2371584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1165446882556771e-05,
+      "loss": 3.3708,
+      "step": 2372096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1157060935046251e-05,
+      "loss": 3.3466,
+      "step": 2372608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1148674987535731e-05,
+      "loss": 3.3602,
+      "step": 2373120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1140289040025211e-05,
+      "loss": 3.351,
+      "step": 2373632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1131919471318422e-05,
+      "loss": 3.3577,
+      "step": 2374144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1123533523807902e-05,
+      "loss": 3.3593,
+      "step": 2374656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1115147576297382e-05,
+      "loss": 3.3658,
+      "step": 2375168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1106761628786862e-05,
+      "loss": 3.3575,
+      "step": 2375680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1098392060080074e-05,
+      "loss": 3.3801,
+      "step": 2376192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1090006112569554e-05,
+      "loss": 3.366,
+      "step": 2376704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1081620165059032e-05,
+      "loss": 3.3621,
+      "step": 2377216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1073234217548512e-05,
+      "loss": 3.3559,
+      "step": 2377728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1064864648841725e-05,
+      "loss": 3.3735,
+      "step": 2378240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1056478701331203e-05,
+      "loss": 3.3612,
+      "step": 2378752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1048092753820683e-05,
+      "loss": 3.357,
+      "step": 2379264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1039706806310165e-05,
+      "loss": 3.3645,
+      "step": 2379776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1031337237603375e-05,
+      "loss": 3.3557,
+      "step": 2380288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1022951290092855e-05,
+      "loss": 3.3549,
+      "step": 2380800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1014565342582335e-05,
+      "loss": 3.3641,
+      "step": 2381312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1006179395071815e-05,
+      "loss": 3.3607,
+      "step": 2381824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0997809826365028e-05,
+      "loss": 3.3643,
+      "step": 2382336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0989423878854506e-05,
+      "loss": 3.3649,
+      "step": 2382848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0981037931343986e-05,
+      "loss": 3.3701,
+      "step": 2383360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0972651983833466e-05,
+      "loss": 3.3598,
+      "step": 2383872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0964282415126677e-05,
+      "loss": 3.3655,
+      "step": 2384384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0955896467616157e-05,
+      "loss": 3.35,
+      "step": 2384896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0947510520105637e-05,
+      "loss": 3.3601,
+      "step": 2385408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0939124572595118e-05,
+      "loss": 3.3561,
+      "step": 2385920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0930755003888327e-05,
+      "loss": 3.3524,
+      "step": 2386432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0922369056377809e-05,
+      "loss": 3.3578,
+      "step": 2386944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0913983108867289e-05,
+      "loss": 3.3678,
+      "step": 2387456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0905597161356767e-05,
+      "loss": 3.3537,
+      "step": 2387968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.089722759264998e-05,
+      "loss": 3.3631,
+      "step": 2388480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.088884164513946e-05,
+      "loss": 3.3647,
+      "step": 2388992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.088045569762894e-05,
+      "loss": 3.3688,
+      "step": 2389504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0872069750118418e-05,
+      "loss": 3.3605,
+      "step": 2390016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.086370018141163e-05,
+      "loss": 3.3411,
+      "step": 2390528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.085531423390111e-05,
+      "loss": 3.36,
+      "step": 2391040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.084692828639059e-05,
+      "loss": 3.3666,
+      "step": 2391552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0838558717683801e-05,
+      "loss": 3.3637,
+      "step": 2392064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0830172770173281e-05,
+      "loss": 3.3527,
+      "step": 2392576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0821786822662763e-05,
+      "loss": 3.3536,
+      "step": 2393088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0813417253955974e-05,
+      "loss": 3.347,
+      "step": 2393600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0805031306445452e-05,
+      "loss": 3.347,
+      "step": 2394112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0796645358934933e-05,
+      "loss": 3.3623,
+      "step": 2394624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0788259411424413e-05,
+      "loss": 3.3467,
+      "step": 2395136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0779873463913892e-05,
+      "loss": 3.3658,
+      "step": 2395648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0771487516403372e-05,
+      "loss": 3.3547,
+      "step": 2396160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0763101568892853e-05,
+      "loss": 3.3508,
+      "step": 2396672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0754732000186064e-05,
+      "loss": 3.3671,
+      "step": 2397184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0746346052675544e-05,
+      "loss": 3.3454,
+      "step": 2397696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0737960105165024e-05,
+      "loss": 3.3519,
+      "step": 2398208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0729574157654504e-05,
+      "loss": 3.349,
+      "step": 2398720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0721204588947715e-05,
+      "loss": 3.3613,
+      "step": 2399232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0712818641437195e-05,
+      "loss": 3.3463,
+      "step": 2399744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0704432693926675e-05,
+      "loss": 3.3591,
+      "step": 2400256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0696046746416155e-05,
+      "loss": 3.3398,
+      "step": 2400768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0687677177709365e-05,
+      "loss": 3.3537,
+      "step": 2401280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0679291230198845e-05,
+      "loss": 3.3635,
+      "step": 2401792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0670905282688325e-05,
+      "loss": 3.3613,
+      "step": 2402304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0662519335177805e-05,
+      "loss": 3.3593,
+      "step": 2402816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0654149766471016e-05,
+      "loss": 3.3623,
+      "step": 2403328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0645763818960498e-05,
+      "loss": 3.3706,
+      "step": 2403840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0637377871449978e-05,
+      "loss": 3.3548,
+      "step": 2404352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0628991923939456e-05,
+      "loss": 3.3691,
+      "step": 2404864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0620622355232668e-05,
+      "loss": 3.3495,
+      "step": 2405376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0612236407722148e-05,
+      "loss": 3.3546,
+      "step": 2405888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.060386683901536e-05,
+      "loss": 3.3611,
+      "step": 2406400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0595480891504839e-05,
+      "loss": 3.3661,
+      "step": 2406912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0587094943994319e-05,
+      "loss": 3.3574,
+      "step": 2407424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0578708996483799e-05,
+      "loss": 3.3655,
+      "step": 2407936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0570323048973279e-05,
+      "loss": 3.3497,
+      "step": 2408448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0561937101462759e-05,
+      "loss": 3.3392,
+      "step": 2408960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0553551153952239e-05,
+      "loss": 3.3491,
+      "step": 2409472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0545165206441719e-05,
+      "loss": 3.356,
+      "step": 2409984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.053679563773493e-05,
+      "loss": 3.3512,
+      "step": 2410496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.052840969022441e-05,
+      "loss": 3.3516,
+      "step": 2411008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.052002374271389e-05,
+      "loss": 3.3593,
+      "step": 2411520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.051163779520337e-05,
+      "loss": 3.3524,
+      "step": 2412032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.050326822649658e-05,
+      "loss": 3.3528,
+      "step": 2412544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.049488227898606e-05,
+      "loss": 3.3527,
+      "step": 2413056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0486496331475542e-05,
+      "loss": 3.3652,
+      "step": 2413568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.047811038396502e-05,
+      "loss": 3.3656,
+      "step": 2414080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0469740815258233e-05,
+      "loss": 3.3641,
+      "step": 2414592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0461354867747713e-05,
+      "loss": 3.347,
+      "step": 2415104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0452968920237191e-05,
+      "loss": 3.3581,
+      "step": 2415616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.044458297272667e-05,
+      "loss": 3.3498,
+      "step": 2416128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0436213404019883e-05,
+      "loss": 3.3606,
+      "step": 2416640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0427827456509363e-05,
+      "loss": 3.3606,
+      "step": 2417152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0419441508998842e-05,
+      "loss": 3.3546,
+      "step": 2417664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0411055561488323e-05,
+      "loss": 3.3545,
+      "step": 2418176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0402685992781534e-05,
+      "loss": 3.361,
+      "step": 2418688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0394300045271014e-05,
+      "loss": 3.3578,
+      "step": 2419200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0385914097760494e-05,
+      "loss": 3.3531,
+      "step": 2419712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0377528150249974e-05,
+      "loss": 3.3546,
+      "step": 2420224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0369158581543186e-05,
+      "loss": 3.3471,
+      "step": 2420736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0360772634032665e-05,
+      "loss": 3.3549,
+      "step": 2421248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0352386686522145e-05,
+      "loss": 3.3463,
+      "step": 2421760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0344017117815357e-05,
+      "loss": 3.3517,
+      "step": 2422272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0335631170304837e-05,
+      "loss": 3.3492,
+      "step": 2422784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0327245222794315e-05,
+      "loss": 3.3593,
+      "step": 2423296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0318859275283795e-05,
+      "loss": 3.3533,
+      "step": 2423808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0310489706577008e-05,
+      "loss": 3.3538,
+      "step": 2424320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0302103759066488e-05,
+      "loss": 3.3709,
+      "step": 2424832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0293717811555968e-05,
+      "loss": 3.3655,
+      "step": 2425344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0285331864045448e-05,
+      "loss": 3.3618,
+      "step": 2425856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0276962295338658e-05,
+      "loss": 3.3544,
+      "step": 2426368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.026859272663187e-05,
+      "loss": 3.3556,
+      "step": 2426880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0260206779121349e-05,
+      "loss": 3.3508,
+      "step": 2427392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0251820831610829e-05,
+      "loss": 3.346,
+      "step": 2427904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.024343488410031e-05,
+      "loss": 3.3549,
+      "step": 2428416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0235048936589789e-05,
+      "loss": 3.3529,
+      "step": 2428928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0226662989079269e-05,
+      "loss": 3.3596,
+      "step": 2429440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0218277041568749e-05,
+      "loss": 3.3545,
+      "step": 2429952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0209891094058229e-05,
+      "loss": 3.3592,
+      "step": 2430464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.020152152535144e-05,
+      "loss": 3.3528,
+      "step": 2430976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0193135577840921e-05,
+      "loss": 3.3515,
+      "step": 2431488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0184749630330401e-05,
+      "loss": 3.3553,
+      "step": 2432000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.017636368281988e-05,
+      "loss": 3.3612,
+      "step": 2432512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0167994114113092e-05,
+      "loss": 3.3681,
+      "step": 2433024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0159608166602572e-05,
+      "loss": 3.3576,
+      "step": 2433536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0151222219092052e-05,
+      "loss": 3.3603,
+      "step": 2434048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.014283627158153e-05,
+      "loss": 3.3573,
+      "step": 2434560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0134466702874743e-05,
+      "loss": 3.3595,
+      "step": 2435072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0126080755364223e-05,
+      "loss": 3.3557,
+      "step": 2435584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0117694807853703e-05,
+      "loss": 3.3494,
+      "step": 2436096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0109308860343182e-05,
+      "loss": 3.3621,
+      "step": 2436608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0100939291636393e-05,
+      "loss": 3.3601,
+      "step": 2437120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0092553344125873e-05,
+      "loss": 3.3539,
+      "step": 2437632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0084167396615353e-05,
+      "loss": 3.3541,
+      "step": 2438144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0075781449104833e-05,
+      "loss": 3.3579,
+      "step": 2438656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0067411880398046e-05,
+      "loss": 3.3605,
+      "step": 2439168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0059025932887526e-05,
+      "loss": 3.35,
+      "step": 2439680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0050639985377004e-05,
+      "loss": 3.3622,
+      "step": 2440192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0042254037866484e-05,
+      "loss": 3.3529,
+      "step": 2440704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0033884469159696e-05,
+      "loss": 3.3694,
+      "step": 2441216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0025498521649175e-05,
+      "loss": 3.3544,
+      "step": 2441728
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0017112574138656e-05,
+      "loss": 3.3521,
+      "step": 2442240
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8596630096435547,
+      "eval_runtime": 303.1844,
+      "eval_samples_per_second": 1258.61,
+      "eval_steps_per_second": 39.332,
+      "step": 2442240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0008726626628136e-05,
+      "loss": 3.3472,
+      "step": 2442752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000340679117616e-05,
+      "loss": 3.3508,
+      "step": 2443264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.991971110410827e-06,
+      "loss": 3.3675,
+      "step": 2443776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.983585162900307e-06,
+      "loss": 3.3617,
+      "step": 2444288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.975199215389787e-06,
+      "loss": 3.3585,
+      "step": 2444800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.966813267879265e-06,
+      "loss": 3.3523,
+      "step": 2445312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.958427320368747e-06,
+      "loss": 3.3548,
+      "step": 2445824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.950041372858227e-06,
+      "loss": 3.3402,
+      "step": 2446336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.941655425347707e-06,
+      "loss": 3.3577,
+      "step": 2446848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.933269477837185e-06,
+      "loss": 3.3495,
+      "step": 2447360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.924899909130397e-06,
+      "loss": 3.3707,
+      "step": 2447872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.916513961619877e-06,
+      "loss": 3.3646,
+      "step": 2448384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.908128014109357e-06,
+      "loss": 3.3376,
+      "step": 2448896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.899742066598837e-06,
+      "loss": 3.3485,
+      "step": 2449408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.891372497892048e-06,
+      "loss": 3.3459,
+      "step": 2449920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.88298655038153e-06,
+      "loss": 3.3454,
+      "step": 2450432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.874600602871008e-06,
+      "loss": 3.355,
+      "step": 2450944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.866214655360488e-06,
+      "loss": 3.3564,
+      "step": 2451456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.8578450866537e-06,
+      "loss": 3.3457,
+      "step": 2451968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.84945913914318e-06,
+      "loss": 3.3682,
+      "step": 2452480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.841073191632659e-06,
+      "loss": 3.3535,
+      "step": 2452992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.832687244122139e-06,
+      "loss": 3.3532,
+      "step": 2453504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.824317675415351e-06,
+      "loss": 3.3487,
+      "step": 2454016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.81593172790483e-06,
+      "loss": 3.366,
+      "step": 2454528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.807545780394311e-06,
+      "loss": 3.3494,
+      "step": 2455040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.799159832883791e-06,
+      "loss": 3.3482,
+      "step": 2455552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.790790264177002e-06,
+      "loss": 3.3531,
+      "step": 2456064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.782404316666482e-06,
+      "loss": 3.3475,
+      "step": 2456576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.774018369155962e-06,
+      "loss": 3.3447,
+      "step": 2457088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.765648800449172e-06,
+      "loss": 3.3513,
+      "step": 2457600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.757262852938652e-06,
+      "loss": 3.3522,
+      "step": 2458112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.748876905428132e-06,
+      "loss": 3.3549,
+      "step": 2458624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.740490957917612e-06,
+      "loss": 3.3536,
+      "step": 2459136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.732121389210825e-06,
+      "loss": 3.3597,
+      "step": 2459648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.723735441700303e-06,
+      "loss": 3.353,
+      "step": 2460160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.715349494189783e-06,
+      "loss": 3.3565,
+      "step": 2460672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.706963546679265e-06,
+      "loss": 3.3424,
+      "step": 2461184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.698593977972475e-06,
+      "loss": 3.3512,
+      "step": 2461696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.690208030461954e-06,
+      "loss": 3.3419,
+      "step": 2462208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.681822082951435e-06,
+      "loss": 3.3474,
+      "step": 2462720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.673436135440915e-06,
+      "loss": 3.3461,
+      "step": 2463232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.665066566734126e-06,
+      "loss": 3.3592,
+      "step": 2463744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.656680619223606e-06,
+      "loss": 3.3469,
+      "step": 2464256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.648294671713086e-06,
+      "loss": 3.3554,
+      "step": 2464768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.639908724202566e-06,
+      "loss": 3.353,
+      "step": 2465280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.631539155495777e-06,
+      "loss": 3.3638,
+      "step": 2465792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.623153207985257e-06,
+      "loss": 3.3497,
+      "step": 2466304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.614767260474737e-06,
+      "loss": 3.3356,
+      "step": 2466816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.606381312964217e-06,
+      "loss": 3.344,
+      "step": 2467328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.598011744257427e-06,
+      "loss": 3.3614,
+      "step": 2467840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.589625796746907e-06,
+      "loss": 3.3517,
+      "step": 2468352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.581239849236389e-06,
+      "loss": 3.3414,
+      "step": 2468864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.572853901725867e-06,
+      "loss": 3.3396,
+      "step": 2469376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.56448433301908e-06,
+      "loss": 3.3451,
+      "step": 2469888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.55609838550856e-06,
+      "loss": 3.3349,
+      "step": 2470400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.54771243799804e-06,
+      "loss": 3.3558,
+      "step": 2470912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.539326490487518e-06,
+      "loss": 3.3388,
+      "step": 2471424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.530940542977e-06,
+      "loss": 3.3493,
+      "step": 2471936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.52255459546648e-06,
+      "loss": 3.348,
+      "step": 2472448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.514168647955958e-06,
+      "loss": 3.3384,
+      "step": 2472960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.50579907924917e-06,
+      "loss": 3.3565,
+      "step": 2473472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.49741313173865e-06,
+      "loss": 3.3396,
+      "step": 2473984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.48902718422813e-06,
+      "loss": 3.3467,
+      "step": 2474496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.480641236717608e-06,
+      "loss": 3.3337,
+      "step": 2475008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.472271668010821e-06,
+      "loss": 3.3552,
+      "step": 2475520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.463885720500301e-06,
+      "loss": 3.3413,
+      "step": 2476032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.45549977298978e-06,
+      "loss": 3.3433,
+      "step": 2476544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.44711382547926e-06,
+      "loss": 3.3357,
+      "step": 2477056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.438744256772472e-06,
+      "loss": 3.3424,
+      "step": 2477568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.430358309261953e-06,
+      "loss": 3.3519,
+      "step": 2478080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.421972361751431e-06,
+      "loss": 3.3558,
+      "step": 2478592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.413586414240911e-06,
+      "loss": 3.3497,
+      "step": 2479104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.405216845534124e-06,
+      "loss": 3.3502,
+      "step": 2479616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.396830898023604e-06,
+      "loss": 3.3571,
+      "step": 2480128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.388444950513082e-06,
+      "loss": 3.345,
+      "step": 2480640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.380059003002562e-06,
+      "loss": 3.3626,
+      "step": 2481152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.371689434295775e-06,
+      "loss": 3.3411,
+      "step": 2481664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.363303486785253e-06,
+      "loss": 3.3425,
+      "step": 2482176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.354917539274734e-06,
+      "loss": 3.3503,
+      "step": 2482688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.346531591764214e-06,
+      "loss": 3.3544,
+      "step": 2483200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.338162023057425e-06,
+      "loss": 3.3488,
+      "step": 2483712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.329776075546905e-06,
+      "loss": 3.3555,
+      "step": 2484224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.321390128036385e-06,
+      "loss": 3.3434,
+      "step": 2484736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.313004180525865e-06,
+      "loss": 3.3271,
+      "step": 2485248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.304634611819078e-06,
+      "loss": 3.3383,
+      "step": 2485760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.296248664308556e-06,
+      "loss": 3.343,
+      "step": 2486272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.287862716798036e-06,
+      "loss": 3.3427,
+      "step": 2486784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.279476769287516e-06,
+      "loss": 3.343,
+      "step": 2487296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.271107200580727e-06,
+      "loss": 3.3525,
+      "step": 2487808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.262721253070207e-06,
+      "loss": 3.3408,
+      "step": 2488320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.254335305559688e-06,
+      "loss": 3.3457,
+      "step": 2488832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.245949358049168e-06,
+      "loss": 3.3443,
+      "step": 2489344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.237579789342377e-06,
+      "loss": 3.3528,
+      "step": 2489856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.229193841831859e-06,
+      "loss": 3.3558,
+      "step": 2490368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.220807894321339e-06,
+      "loss": 3.3588,
+      "step": 2490880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.212421946810817e-06,
+      "loss": 3.3332,
+      "step": 2491392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.20405237810403e-06,
+      "loss": 3.3476,
+      "step": 2491904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.19566643059351e-06,
+      "loss": 3.3429,
+      "step": 2492416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.18728048308299e-06,
+      "loss": 3.3482,
+      "step": 2492928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.17889453557247e-06,
+      "loss": 3.352,
+      "step": 2493440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.17052496686568e-06,
+      "loss": 3.3449,
+      "step": 2493952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.16213901935516e-06,
+      "loss": 3.343,
+      "step": 2494464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.15375307184464e-06,
+      "loss": 3.3515,
+      "step": 2494976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.14536712433412e-06,
+      "loss": 3.3486,
+      "step": 2495488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.136997555627331e-06,
+      "loss": 3.3411,
+      "step": 2496000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.128611608116813e-06,
+      "loss": 3.349,
+      "step": 2496512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.12022566060629e-06,
+      "loss": 3.3385,
+      "step": 2497024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.11183971309577e-06,
+      "loss": 3.3413,
+      "step": 2497536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.103470144388983e-06,
+      "loss": 3.3363,
+      "step": 2498048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.095084196878463e-06,
+      "loss": 3.3409,
+      "step": 2498560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.086698249367941e-06,
+      "loss": 3.3398,
+      "step": 2499072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.078312301857423e-06,
+      "loss": 3.3511,
+      "step": 2499584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.069942733150634e-06,
+      "loss": 3.343,
+      "step": 2500096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.061556785640112e-06,
+      "loss": 3.3445,
+      "step": 2500608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.053170838129594e-06,
+      "loss": 3.3629,
+      "step": 2501120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.044784890619074e-06,
+      "loss": 3.3571,
+      "step": 2501632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.036415321912285e-06,
+      "loss": 3.3476,
+      "step": 2502144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.028029374401764e-06,
+      "loss": 3.3485,
+      "step": 2502656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.019659805694975e-06,
+      "loss": 3.3452,
+      "step": 2503168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.011273858184455e-06,
+      "loss": 3.3395,
+      "step": 2503680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.002887910673937e-06,
+      "loss": 3.3373,
+      "step": 2504192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.994501963163415e-06,
+      "loss": 3.3438,
+      "step": 2504704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.986116015652895e-06,
+      "loss": 3.3463,
+      "step": 2505216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.977730068142377e-06,
+      "loss": 3.3505,
+      "step": 2505728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.969344120631855e-06,
+      "loss": 3.3452,
+      "step": 2506240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.960958173121335e-06,
+      "loss": 3.3492,
+      "step": 2506752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.952588604414547e-06,
+      "loss": 3.3441,
+      "step": 2507264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.944202656904027e-06,
+      "loss": 3.3424,
+      "step": 2507776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.935816709393506e-06,
+      "loss": 3.3405,
+      "step": 2508288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.927430761882986e-06,
+      "loss": 3.3569,
+      "step": 2508800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.919061193176198e-06,
+      "loss": 3.3588,
+      "step": 2509312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.910675245665676e-06,
+      "loss": 3.3458,
+      "step": 2509824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.902289298155158e-06,
+      "loss": 3.3486,
+      "step": 2510336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.893903350644638e-06,
+      "loss": 3.3496,
+      "step": 2510848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.885533781937849e-06,
+      "loss": 3.3488,
+      "step": 2511360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.877147834427329e-06,
+      "loss": 3.3505,
+      "step": 2511872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.868761886916809e-06,
+      "loss": 3.3421,
+      "step": 2512384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.860375939406289e-06,
+      "loss": 3.3536,
+      "step": 2512896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.852006370699501e-06,
+      "loss": 3.3458,
+      "step": 2513408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.84362042318898e-06,
+      "loss": 3.3462,
+      "step": 2513920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.83523447567846e-06,
+      "loss": 3.3478,
+      "step": 2514432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.82684852816794e-06,
+      "loss": 3.3461,
+      "step": 2514944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.81847895946115e-06,
+      "loss": 3.3504,
+      "step": 2515456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.81009301195063e-06,
+      "loss": 3.3407,
+      "step": 2515968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.801707064440112e-06,
+      "loss": 3.3539,
+      "step": 2516480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.793321116929592e-06,
+      "loss": 3.3459,
+      "step": 2516992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.7849515482228e-06,
+      "loss": 3.3555,
+      "step": 2517504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.776565600712282e-06,
+      "loss": 3.3483,
+      "step": 2518016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.768179653201762e-06,
+      "loss": 3.3394,
+      "step": 2518528
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8602757453918457,
+      "eval_runtime": 320.7307,
+      "eval_samples_per_second": 1189.755,
+      "eval_steps_per_second": 37.181,
+      "step": 2518560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.75979370569124e-06,
+      "loss": 3.3394,
+      "step": 2519040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.751424136984453e-06,
+      "loss": 3.3431,
+      "step": 2519552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.743038189473933e-06,
+      "loss": 3.3618,
+      "step": 2520064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.734652241963413e-06,
+      "loss": 3.3491,
+      "step": 2520576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.726266294452893e-06,
+      "loss": 3.3531,
+      "step": 2521088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.717896725746104e-06,
+      "loss": 3.3452,
+      "step": 2521600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.709510778235584e-06,
+      "loss": 3.3426,
+      "step": 2522112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.701124830725064e-06,
+      "loss": 3.3345,
+      "step": 2522624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.692738883214544e-06,
+      "loss": 3.347,
+      "step": 2523136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.684369314507754e-06,
+      "loss": 3.3403,
+      "step": 2523648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.675983366997236e-06,
+      "loss": 3.3609,
+      "step": 2524160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.667597419486714e-06,
+      "loss": 3.3503,
+      "step": 2524672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.659227850779927e-06,
+      "loss": 3.3309,
+      "step": 2525184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.650841903269407e-06,
+      "loss": 3.3389,
+      "step": 2525696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.642455955758887e-06,
+      "loss": 3.3407,
+      "step": 2526208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.634070008248365e-06,
+      "loss": 3.3277,
+      "step": 2526720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.625700439541577e-06,
+      "loss": 3.3477,
+      "step": 2527232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.617314492031057e-06,
+      "loss": 3.345,
+      "step": 2527744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.608928544520537e-06,
+      "loss": 3.3362,
+      "step": 2528256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.600542597010017e-06,
+      "loss": 3.3615,
+      "step": 2528768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.592173028303228e-06,
+      "loss": 3.3454,
+      "step": 2529280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.583787080792708e-06,
+      "loss": 3.3466,
+      "step": 2529792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.575401133282188e-06,
+      "loss": 3.3402,
+      "step": 2530304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.567015185771668e-06,
+      "loss": 3.3561,
+      "step": 2530816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.558645617064879e-06,
+      "loss": 3.3355,
+      "step": 2531328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.55025966955436e-06,
+      "loss": 3.3439,
+      "step": 2531840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.541873722043839e-06,
+      "loss": 3.3415,
+      "step": 2532352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.533487774533319e-06,
+      "loss": 3.3425,
+      "step": 2532864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.525118205826531e-06,
+      "loss": 3.3345,
+      "step": 2533376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.51673225831601e-06,
+      "loss": 3.3453,
+      "step": 2533888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.50834631080549e-06,
+      "loss": 3.3444,
+      "step": 2534400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.499960363294971e-06,
+      "loss": 3.3445,
+      "step": 2534912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.491590794588182e-06,
+      "loss": 3.3456,
+      "step": 2535424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.483204847077662e-06,
+      "loss": 3.3538,
+      "step": 2535936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.474818899567142e-06,
+      "loss": 3.3427,
+      "step": 2536448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.466432952056622e-06,
+      "loss": 3.3467,
+      "step": 2536960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.458063383349832e-06,
+      "loss": 3.3353,
+      "step": 2537472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.449677435839312e-06,
+      "loss": 3.3418,
+      "step": 2537984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.441291488328792e-06,
+      "loss": 3.3312,
+      "step": 2538496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.432905540818272e-06,
+      "loss": 3.3386,
+      "step": 2539008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.424535972111483e-06,
+      "loss": 3.3372,
+      "step": 2539520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.416150024600963e-06,
+      "loss": 3.3526,
+      "step": 2540032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.407764077090443e-06,
+      "loss": 3.3345,
+      "step": 2540544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.399378129579925e-06,
+      "loss": 3.3485,
+      "step": 2541056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.391008560873134e-06,
+      "loss": 3.343,
+      "step": 2541568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.382622613362614e-06,
+      "loss": 3.3523,
+      "step": 2542080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.374236665852095e-06,
+      "loss": 3.3401,
+      "step": 2542592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.365850718341574e-06,
+      "loss": 3.3281,
+      "step": 2543104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.357481149634786e-06,
+      "loss": 3.3369,
+      "step": 2543616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.349095202124266e-06,
+      "loss": 3.3494,
+      "step": 2544128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.340709254613746e-06,
+      "loss": 3.3421,
+      "step": 2544640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.332323307103224e-06,
+      "loss": 3.3349,
+      "step": 2545152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.323953738396437e-06,
+      "loss": 3.3287,
+      "step": 2545664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.315584169689648e-06,
+      "loss": 3.3386,
+      "step": 2546176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.30719822217913e-06,
+      "loss": 3.3224,
+      "step": 2546688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.298812274668607e-06,
+      "loss": 3.3504,
+      "step": 2547200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.290426327158087e-06,
+      "loss": 3.3297,
+      "step": 2547712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.282040379647567e-06,
+      "loss": 3.3438,
+      "step": 2548224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.273654432137047e-06,
+      "loss": 3.3373,
+      "step": 2548736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.265268484626527e-06,
+      "loss": 3.33,
+      "step": 2549248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.256882537116007e-06,
+      "loss": 3.3435,
+      "step": 2549760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.24851296840922e-06,
+      "loss": 3.3318,
+      "step": 2550272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.240127020898698e-06,
+      "loss": 3.3368,
+      "step": 2550784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.231741073388178e-06,
+      "loss": 3.3242,
+      "step": 2551296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.22337150468139e-06,
+      "loss": 3.3473,
+      "step": 2551808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.21498555717087e-06,
+      "loss": 3.3299,
+      "step": 2552320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.206615988464081e-06,
+      "loss": 3.336,
+      "step": 2552832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.198230040953561e-06,
+      "loss": 3.3245,
+      "step": 2553344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.189844093443041e-06,
+      "loss": 3.3363,
+      "step": 2553856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.181458145932521e-06,
+      "loss": 3.3402,
+      "step": 2554368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.173072198422001e-06,
+      "loss": 3.3456,
+      "step": 2554880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.164702629715212e-06,
+      "loss": 3.3404,
+      "step": 2555392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.156316682204693e-06,
+      "loss": 3.3445,
+      "step": 2555904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.147930734694172e-06,
+      "loss": 3.3483,
+      "step": 2556416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.139544787183652e-06,
+      "loss": 3.341,
+      "step": 2556928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.131158839673132e-06,
+      "loss": 3.3491,
+      "step": 2557440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.122772892162612e-06,
+      "loss": 3.3377,
+      "step": 2557952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.114386944652092e-06,
+      "loss": 3.3355,
+      "step": 2558464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.106017375945302e-06,
+      "loss": 3.3392,
+      "step": 2558976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.097631428434784e-06,
+      "loss": 3.3452,
+      "step": 2559488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.089245480924262e-06,
+      "loss": 3.3416,
+      "step": 2560000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.080859533413742e-06,
+      "loss": 3.348,
+      "step": 2560512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.072473585903224e-06,
+      "loss": 3.338,
+      "step": 2561024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.064087638392702e-06,
+      "loss": 3.3177,
+      "step": 2561536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.055701690882182e-06,
+      "loss": 3.3281,
+      "step": 2562048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.047315743371662e-06,
+      "loss": 3.3345,
+      "step": 2562560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.038946174664875e-06,
+      "loss": 3.3366,
+      "step": 2563072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.030560227154353e-06,
+      "loss": 3.3321,
+      "step": 2563584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.022174279643833e-06,
+      "loss": 3.3455,
+      "step": 2564096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.013788332133314e-06,
+      "loss": 3.3294,
+      "step": 2564608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.005418763426525e-06,
+      "loss": 3.332,
+      "step": 2565120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.997032815916005e-06,
+      "loss": 3.3349,
+      "step": 2565632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.988646868405485e-06,
+      "loss": 3.3465,
+      "step": 2566144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.980260920894965e-06,
+      "loss": 3.3443,
+      "step": 2566656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.971891352188176e-06,
+      "loss": 3.3522,
+      "step": 2567168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.963505404677656e-06,
+      "loss": 3.326,
+      "step": 2567680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.955119457167136e-06,
+      "loss": 3.3383,
+      "step": 2568192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.946733509656616e-06,
+      "loss": 3.3389,
+      "step": 2568704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.938363940949826e-06,
+      "loss": 3.3356,
+      "step": 2569216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.929977993439306e-06,
+      "loss": 3.345,
+      "step": 2569728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.921592045928786e-06,
+      "loss": 3.3337,
+      "step": 2570240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.913206098418266e-06,
+      "loss": 3.3355,
+      "step": 2570752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.904836529711477e-06,
+      "loss": 3.3445,
+      "step": 2571264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.896450582200959e-06,
+      "loss": 3.3422,
+      "step": 2571776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.888064634690439e-06,
+      "loss": 3.327,
+      "step": 2572288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.879678687179917e-06,
+      "loss": 3.3408,
+      "step": 2572800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.87130911847313e-06,
+      "loss": 3.3339,
+      "step": 2573312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.86292317096261e-06,
+      "loss": 3.332,
+      "step": 2573824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.854537223452088e-06,
+      "loss": 3.3284,
+      "step": 2574336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.846151275941568e-06,
+      "loss": 3.3322,
+      "step": 2574848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.83778170723478e-06,
+      "loss": 3.3305,
+      "step": 2575360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.82939575972426e-06,
+      "loss": 3.3402,
+      "step": 2575872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.82100981221374e-06,
+      "loss": 3.3311,
+      "step": 2576384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.81262386470322e-06,
+      "loss": 3.3364,
+      "step": 2576896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.80425429599643e-06,
+      "loss": 3.3545,
+      "step": 2577408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.79586834848591e-06,
+      "loss": 3.3459,
+      "step": 2577920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.78748240097539e-06,
+      "loss": 3.3403,
+      "step": 2578432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.779112832268602e-06,
+      "loss": 3.3407,
+      "step": 2578944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.770743263561814e-06,
+      "loss": 3.3347,
+      "step": 2579456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.762357316051294e-06,
+      "loss": 3.3314,
+      "step": 2579968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.753971368540772e-06,
+      "loss": 3.332,
+      "step": 2580480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.745585421030254e-06,
+      "loss": 3.3315,
+      "step": 2580992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.737199473519734e-06,
+      "loss": 3.3389,
+      "step": 2581504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.728813526009212e-06,
+      "loss": 3.3425,
+      "step": 2582016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.720427578498694e-06,
+      "loss": 3.3391,
+      "step": 2582528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.712041630988174e-06,
+      "loss": 3.3375,
+      "step": 2583040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.703672062281384e-06,
+      "loss": 3.3372,
+      "step": 2583552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.695286114770864e-06,
+      "loss": 3.3311,
+      "step": 2584064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.686900167260344e-06,
+      "loss": 3.3322,
+      "step": 2584576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.678514219749824e-06,
+      "loss": 3.3455,
+      "step": 2585088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.670144651043035e-06,
+      "loss": 3.3504,
+      "step": 2585600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.661758703532515e-06,
+      "loss": 3.342,
+      "step": 2586112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.653372756021995e-06,
+      "loss": 3.3352,
+      "step": 2586624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.644986808511475e-06,
+      "loss": 3.34,
+      "step": 2587136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.636617239804686e-06,
+      "loss": 3.338,
+      "step": 2587648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.628231292294167e-06,
+      "loss": 3.3443,
+      "step": 2588160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.6198453447836465e-06,
+      "loss": 3.3365,
+      "step": 2588672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.611459397273126e-06,
+      "loss": 3.3424,
+      "step": 2589184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.603089828566337e-06,
+      "loss": 3.3367,
+      "step": 2589696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.594703881055817e-06,
+      "loss": 3.3401,
+      "step": 2590208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.586317933545298e-06,
+      "loss": 3.3347,
+      "step": 2590720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.577931986034776e-06,
+      "loss": 3.3396,
+      "step": 2591232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.569562417327988e-06,
+      "loss": 3.3451,
+      "step": 2591744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.561176469817469e-06,
+      "loss": 3.3298,
+      "step": 2592256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.552790522306949e-06,
+      "loss": 3.3477,
+      "step": 2592768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.544404574796428e-06,
+      "loss": 3.341,
+      "step": 2593280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.5360350060896395e-06,
+      "loss": 3.3438,
+      "step": 2593792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.52764905857912e-06,
+      "loss": 3.3389,
+      "step": 2594304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.5192631110685985e-06,
+      "loss": 3.331,
+      "step": 2594816
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.860727071762085,
+      "eval_runtime": 319.4099,
+      "eval_samples_per_second": 1194.675,
+      "eval_steps_per_second": 37.334,
+      "step": 2594880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.510877163558079e-06,
+      "loss": 3.3298,
+      "step": 2595328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.502491216047559e-06,
+      "loss": 3.3332,
+      "step": 2595840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.49410526853704e-06,
+      "loss": 3.3505,
+      "step": 2596352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.485719321026518e-06,
+      "loss": 3.3366,
+      "step": 2596864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.47734975231973e-06,
+      "loss": 3.3477,
+      "step": 2597376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.468963804809211e-06,
+      "loss": 3.333,
+      "step": 2597888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.46057785729869e-06,
+      "loss": 3.3359,
+      "step": 2598400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4522082885919015e-06,
+      "loss": 3.3229,
+      "step": 2598912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4438223410813815e-06,
+      "loss": 3.3382,
+      "step": 2599424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4354363935708614e-06,
+      "loss": 3.3337,
+      "step": 2599936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4270504460603405e-06,
+      "loss": 3.3504,
+      "step": 2600448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.418680877353552e-06,
+      "loss": 3.3417,
+      "step": 2600960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.410294929843033e-06,
+      "loss": 3.3248,
+      "step": 2601472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.401908982332513e-06,
+      "loss": 3.327,
+      "step": 2601984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.393523034821992e-06,
+      "loss": 3.3347,
+      "step": 2602496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.385153466115204e-06,
+      "loss": 3.3192,
+      "step": 2603008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.376767518604684e-06,
+      "loss": 3.3344,
+      "step": 2603520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.368381571094163e-06,
+      "loss": 3.3389,
+      "step": 2604032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3599956235836436e-06,
+      "loss": 3.3277,
+      "step": 2604544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.351626054876855e-06,
+      "loss": 3.3547,
+      "step": 2605056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.343240107366335e-06,
+      "loss": 3.3348,
+      "step": 2605568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.334854159855814e-06,
+      "loss": 3.3406,
+      "step": 2606080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.326468212345294e-06,
+      "loss": 3.3314,
+      "step": 2606592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.318098643638506e-06,
+      "loss": 3.3469,
+      "step": 2607104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.309712696127987e-06,
+      "loss": 3.3249,
+      "step": 2607616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.301326748617465e-06,
+      "loss": 3.3358,
+      "step": 2608128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.292940801106946e-06,
+      "loss": 3.3293,
+      "step": 2608640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.284571232400157e-06,
+      "loss": 3.3338,
+      "step": 2609152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.2761852848896365e-06,
+      "loss": 3.3262,
+      "step": 2609664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.267799337379116e-06,
+      "loss": 3.3344,
+      "step": 2610176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.259413389868596e-06,
+      "loss": 3.3376,
+      "step": 2610688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.251043821161808e-06,
+      "loss": 3.3371,
+      "step": 2611200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.242657873651287e-06,
+      "loss": 3.3329,
+      "step": 2611712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.234271926140768e-06,
+      "loss": 3.3462,
+      "step": 2612224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.225885978630248e-06,
+      "loss": 3.3311,
+      "step": 2612736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.217516409923458e-06,
+      "loss": 3.3373,
+      "step": 2613248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.209130462412939e-06,
+      "loss": 3.3279,
+      "step": 2613760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.200744514902419e-06,
+      "loss": 3.332,
+      "step": 2614272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.192358567391899e-06,
+      "loss": 3.3209,
+      "step": 2614784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.183988998685109e-06,
+      "loss": 3.3304,
+      "step": 2615296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.17560305117459e-06,
+      "loss": 3.327,
+      "step": 2615808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.16721710366407e-06,
+      "loss": 3.3422,
+      "step": 2616320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.158831156153549e-06,
+      "loss": 3.3261,
+      "step": 2616832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.150461587446761e-06,
+      "loss": 3.3351,
+      "step": 2617344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.142075639936241e-06,
+      "loss": 3.3397,
+      "step": 2617856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.133689692425722e-06,
+      "loss": 3.3402,
+      "step": 2618368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.1253037449152e-06,
+      "loss": 3.3351,
+      "step": 2618880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.1169341762084115e-06,
+      "loss": 3.3199,
+      "step": 2619392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.108548228697892e-06,
+      "loss": 3.3274,
+      "step": 2619904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.100162281187372e-06,
+      "loss": 3.3415,
+      "step": 2620416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.091776333676851e-06,
+      "loss": 3.3322,
+      "step": 2620928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.083406764970063e-06,
+      "loss": 3.3281,
+      "step": 2621440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.075020817459543e-06,
+      "loss": 3.3186,
+      "step": 2621952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.0666512487527554e-06,
+      "loss": 3.3331,
+      "step": 2622464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.058265301242234e-06,
+      "loss": 3.3083,
+      "step": 2622976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.0498793537317145e-06,
+      "loss": 3.3421,
+      "step": 2623488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.0414934062211945e-06,
+      "loss": 3.3184,
+      "step": 2624000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.0331074587106736e-06,
+      "loss": 3.3336,
+      "step": 2624512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.0247215112001535e-06,
+      "loss": 3.3298,
+      "step": 2625024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.016335563689634e-06,
+      "loss": 3.32,
+      "step": 2625536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.0079496161791134e-06,
+      "loss": 3.3389,
+      "step": 2626048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.999580047472325e-06,
+      "loss": 3.3241,
+      "step": 2626560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.991194099961805e-06,
+      "loss": 3.3257,
+      "step": 2627072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.982808152451285e-06,
+      "loss": 3.3142,
+      "step": 2627584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.974422204940764e-06,
+      "loss": 3.3376,
+      "step": 2628096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.966052636233976e-06,
+      "loss": 3.3232,
+      "step": 2628608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9576666887234565e-06,
+      "loss": 3.3267,
+      "step": 2629120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9492807412129365e-06,
+      "loss": 3.317,
+      "step": 2629632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9409111725061464e-06,
+      "loss": 3.327,
+      "step": 2630144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.932525224995627e-06,
+      "loss": 3.3323,
+      "step": 2630656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.924139277485107e-06,
+      "loss": 3.338,
+      "step": 2631168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.915753329974586e-06,
+      "loss": 3.3271,
+      "step": 2631680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.907383761267798e-06,
+      "loss": 3.3387,
+      "step": 2632192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.898997813757279e-06,
+      "loss": 3.336,
+      "step": 2632704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.890611866246759e-06,
+      "loss": 3.3341,
+      "step": 2633216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.882225918736238e-06,
+      "loss": 3.3349,
+      "step": 2633728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8738563500294494e-06,
+      "loss": 3.3324,
+      "step": 2634240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.865470402518929e-06,
+      "loss": 3.3282,
+      "step": 2634752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.85708445500841e-06,
+      "loss": 3.3291,
+      "step": 2635264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8486985074978885e-06,
+      "loss": 3.3354,
+      "step": 2635776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8403289387911e-06,
+      "loss": 3.3364,
+      "step": 2636288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.831942991280581e-06,
+      "loss": 3.3395,
+      "step": 2636800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.82355704377006e-06,
+      "loss": 3.3255,
+      "step": 2637312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.81517109625954e-06,
+      "loss": 3.3082,
+      "step": 2637824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.806801527552752e-06,
+      "loss": 3.319,
+      "step": 2638336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.7984155800422316e-06,
+      "loss": 3.3257,
+      "step": 2638848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.790029632531711e-06,
+      "loss": 3.3285,
+      "step": 2639360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.7816436850211915e-06,
+      "loss": 3.3214,
+      "step": 2639872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.773274116314403e-06,
+      "loss": 3.3349,
+      "step": 2640384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.764888168803883e-06,
+      "loss": 3.3265,
+      "step": 2640896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.756502221293362e-06,
+      "loss": 3.3161,
+      "step": 2641408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.748116273782842e-06,
+      "loss": 3.3274,
+      "step": 2641920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.739746705076054e-06,
+      "loss": 3.3373,
+      "step": 2642432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.731360757565533e-06,
+      "loss": 3.3337,
+      "step": 2642944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.722974810055014e-06,
+      "loss": 3.3387,
+      "step": 2643456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.714588862544494e-06,
+      "loss": 3.3226,
+      "step": 2643968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.706219293837705e-06,
+      "loss": 3.3287,
+      "step": 2644480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.697833346327184e-06,
+      "loss": 3.3287,
+      "step": 2644992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.689447398816664e-06,
+      "loss": 3.3287,
+      "step": 2645504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.681061451306145e-06,
+      "loss": 3.3345,
+      "step": 2646016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.672691882599357e-06,
+      "loss": 3.3273,
+      "step": 2646528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.664305935088835e-06,
+      "loss": 3.3244,
+      "step": 2647040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.655919987578316e-06,
+      "loss": 3.3345,
+      "step": 2647552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.647534040067796e-06,
+      "loss": 3.3377,
+      "step": 2648064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.639164471361007e-06,
+      "loss": 3.3203,
+      "step": 2648576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.6307785238504865e-06,
+      "loss": 3.3261,
+      "step": 2649088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.6223925763399665e-06,
+      "loss": 3.3282,
+      "step": 2649600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.614006628829447e-06,
+      "loss": 3.323,
+      "step": 2650112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.605637060122657e-06,
+      "loss": 3.3192,
+      "step": 2650624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.597251112612138e-06,
+      "loss": 3.3217,
+      "step": 2651136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.588865165101618e-06,
+      "loss": 3.3279,
+      "step": 2651648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.580479217591097e-06,
+      "loss": 3.3286,
+      "step": 2652160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.572109648884309e-06,
+      "loss": 3.3236,
+      "step": 2652672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.563723701373789e-06,
+      "loss": 3.3281,
+      "step": 2653184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.5553377538632695e-06,
+      "loss": 3.3434,
+      "step": 2653696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.546951806352749e-06,
+      "loss": 3.3403,
+      "step": 2654208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.53858223764596e-06,
+      "loss": 3.3308,
+      "step": 2654720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.53019629013544e-06,
+      "loss": 3.3333,
+      "step": 2655232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.521826721428652e-06,
+      "loss": 3.3272,
+      "step": 2655744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.513440773918131e-06,
+      "loss": 3.3241,
+      "step": 2656256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.505054826407611e-06,
+      "loss": 3.3196,
+      "step": 2656768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.496668878897092e-06,
+      "loss": 3.3215,
+      "step": 2657280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.48828293138657e-06,
+      "loss": 3.3271,
+      "step": 2657792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.479896983876051e-06,
+      "loss": 3.3364,
+      "step": 2658304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.471511036365531e-06,
+      "loss": 3.3293,
+      "step": 2658816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.46312508885501e-06,
+      "loss": 3.3296,
+      "step": 2659328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.4547555201482215e-06,
+      "loss": 3.3264,
+      "step": 2659840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.446369572637702e-06,
+      "loss": 3.3206,
+      "step": 2660352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.437983625127182e-06,
+      "loss": 3.3228,
+      "step": 2660864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.429597677616661e-06,
+      "loss": 3.3379,
+      "step": 2661376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.421228108909873e-06,
+      "loss": 3.3441,
+      "step": 2661888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.412842161399353e-06,
+      "loss": 3.3319,
+      "step": 2662400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.404456213888834e-06,
+      "loss": 3.3265,
+      "step": 2662912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.396070266378312e-06,
+      "loss": 3.3285,
+      "step": 2663424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.387700697671524e-06,
+      "loss": 3.3304,
+      "step": 2663936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3793147501610044e-06,
+      "loss": 3.3398,
+      "step": 2664448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3709288026504835e-06,
+      "loss": 3.3272,
+      "step": 2664960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3625428551399635e-06,
+      "loss": 3.3321,
+      "step": 2665472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.354173286433175e-06,
+      "loss": 3.3271,
+      "step": 2665984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.345787338922655e-06,
+      "loss": 3.3306,
+      "step": 2666496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.337401391412134e-06,
+      "loss": 3.3297,
+      "step": 2667008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.329015443901615e-06,
+      "loss": 3.3283,
+      "step": 2667520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.320645875194827e-06,
+      "loss": 3.3348,
+      "step": 2668032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.312259927684307e-06,
+      "loss": 3.3224,
+      "step": 2668544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.303873980173786e-06,
+      "loss": 3.336,
+      "step": 2669056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.295488032663266e-06,
+      "loss": 3.3364,
+      "step": 2669568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.287118463956477e-06,
+      "loss": 3.333,
+      "step": 2670080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.278732516445956e-06,
+      "loss": 3.3303,
+      "step": 2670592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.270346568935437e-06,
+      "loss": 3.3193,
+      "step": 2671104
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8602688312530518,
+      "eval_runtime": 304.608,
+      "eval_samples_per_second": 1252.728,
+      "eval_steps_per_second": 39.149,
+      "step": 2671200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.261960621424917e-06,
+      "loss": 3.3214,
+      "step": 2671616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.253574673914397e-06,
+      "loss": 3.3239,
+      "step": 2672128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.245205105207609e-06,
+      "loss": 3.3418,
+      "step": 2672640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.236819157697088e-06,
+      "loss": 3.3296,
+      "step": 2673152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.228433210186568e-06,
+      "loss": 3.3441,
+      "step": 2673664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.220047262676048e-06,
+      "loss": 3.3226,
+      "step": 2674176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.211661315165528e-06,
+      "loss": 3.3298,
+      "step": 2674688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.203275367655008e-06,
+      "loss": 3.3122,
+      "step": 2675200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.194889420144488e-06,
+      "loss": 3.3316,
+      "step": 2675712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.186503472633967e-06,
+      "loss": 3.3204,
+      "step": 2676224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.178133903927178e-06,
+      "loss": 3.3411,
+      "step": 2676736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.169747956416658e-06,
+      "loss": 3.3309,
+      "step": 2677248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.161362008906138e-06,
+      "loss": 3.3142,
+      "step": 2677760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.152976061395618e-06,
+      "loss": 3.3189,
+      "step": 2678272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.14460649268883e-06,
+      "loss": 3.3275,
+      "step": 2678784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.13622054517831e-06,
+      "loss": 3.3127,
+      "step": 2679296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.12783459766779e-06,
+      "loss": 3.3249,
+      "step": 2679808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.11944865015727e-06,
+      "loss": 3.3261,
+      "step": 2680320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.111079081450481e-06,
+      "loss": 3.3201,
+      "step": 2680832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.1026931339399605e-06,
+      "loss": 3.3475,
+      "step": 2681344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.0943071864294405e-06,
+      "loss": 3.3282,
+      "step": 2681856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.0859212389189204e-06,
+      "loss": 3.3276,
+      "step": 2682368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.077551670212132e-06,
+      "loss": 3.3272,
+      "step": 2682880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.069165722701612e-06,
+      "loss": 3.3314,
+      "step": 2683392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.060779775191092e-06,
+      "loss": 3.321,
+      "step": 2683904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.052393827680572e-06,
+      "loss": 3.3247,
+      "step": 2684416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.044024258973783e-06,
+      "loss": 3.3212,
+      "step": 2684928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.0356383114632635e-06,
+      "loss": 3.3245,
+      "step": 2685440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.027252363952743e-06,
+      "loss": 3.3166,
+      "step": 2685952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.018866416442223e-06,
+      "loss": 3.3275,
+      "step": 2686464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.010496847735434e-06,
+      "loss": 3.3278,
+      "step": 2686976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.002110900224913e-06,
+      "loss": 3.3281,
+      "step": 2687488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.993724952714394e-06,
+      "loss": 3.3287,
+      "step": 2688000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.985339005203874e-06,
+      "loss": 3.3355,
+      "step": 2688512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.976969436497086e-06,
+      "loss": 3.3213,
+      "step": 2689024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.968583488986565e-06,
+      "loss": 3.3264,
+      "step": 2689536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.960197541476046e-06,
+      "loss": 3.3133,
+      "step": 2690048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.951811593965525e-06,
+      "loss": 3.3286,
+      "step": 2690560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.943442025258736e-06,
+      "loss": 3.3138,
+      "step": 2691072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.935056077748216e-06,
+      "loss": 3.319,
+      "step": 2691584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9266701302376955e-06,
+      "loss": 3.3172,
+      "step": 2692096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.918300561530907e-06,
+      "loss": 3.3355,
+      "step": 2692608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.909914614020387e-06,
+      "loss": 3.3167,
+      "step": 2693120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.901528666509867e-06,
+      "loss": 3.3235,
+      "step": 2693632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.893142718999347e-06,
+      "loss": 3.3345,
+      "step": 2694144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.884773150292559e-06,
+      "loss": 3.3313,
+      "step": 2694656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.8763872027820385e-06,
+      "loss": 3.3294,
+      "step": 2695168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.8680012552715185e-06,
+      "loss": 3.3123,
+      "step": 2695680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.8596153077609985e-06,
+      "loss": 3.3141,
+      "step": 2696192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.85124573905421e-06,
+      "loss": 3.3295,
+      "step": 2696704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.842859791543689e-06,
+      "loss": 3.3259,
+      "step": 2697216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.834473844033169e-06,
+      "loss": 3.3167,
+      "step": 2697728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.826087896522649e-06,
+      "loss": 3.3126,
+      "step": 2698240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.817718327815861e-06,
+      "loss": 3.3264,
+      "step": 2698752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.809332380305341e-06,
+      "loss": 3.3041,
+      "step": 2699264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.800946432794821e-06,
+      "loss": 3.3252,
+      "step": 2699776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.792560485284301e-06,
+      "loss": 3.3126,
+      "step": 2700288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.784174537773781e-06,
+      "loss": 3.3215,
+      "step": 2700800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.77578859026326e-06,
+      "loss": 3.3236,
+      "step": 2701312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.7674026427527405e-06,
+      "loss": 3.3119,
+      "step": 2701824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.75901669524222e-06,
+      "loss": 3.3265,
+      "step": 2702336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.750647126535431e-06,
+      "loss": 3.3168,
+      "step": 2702848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.742261179024911e-06,
+      "loss": 3.3161,
+      "step": 2703360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.733875231514391e-06,
+      "loss": 3.3093,
+      "step": 2703872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.725489284003871e-06,
+      "loss": 3.3252,
+      "step": 2704384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.717119715297083e-06,
+      "loss": 3.317,
+      "step": 2704896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.708733767786562e-06,
+      "loss": 3.3196,
+      "step": 2705408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.700347820276042e-06,
+      "loss": 3.3089,
+      "step": 2705920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.691961872765522e-06,
+      "loss": 3.3138,
+      "step": 2706432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.683592304058733e-06,
+      "loss": 3.3235,
+      "step": 2706944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.675206356548213e-06,
+      "loss": 3.3298,
+      "step": 2707456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.666820409037693e-06,
+      "loss": 3.3193,
+      "step": 2707968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.658450840330905e-06,
+      "loss": 3.3308,
+      "step": 2708480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.650064892820384e-06,
+      "loss": 3.3259,
+      "step": 2708992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.641678945309864e-06,
+      "loss": 3.3221,
+      "step": 2709504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.633292997799344e-06,
+      "loss": 3.3277,
+      "step": 2710016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.624923429092556e-06,
+      "loss": 3.3214,
+      "step": 2710528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.6165374815820356e-06,
+      "loss": 3.3203,
+      "step": 2711040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.608167912875246e-06,
+      "loss": 3.322,
+      "step": 2711552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.599781965364727e-06,
+      "loss": 3.3273,
+      "step": 2712064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.591396017854206e-06,
+      "loss": 3.3279,
+      "step": 2712576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.583010070343687e-06,
+      "loss": 3.3238,
+      "step": 2713088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.574624122833166e-06,
+      "loss": 3.322,
+      "step": 2713600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.566238175322646e-06,
+      "loss": 3.3029,
+      "step": 2714112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.557852227812126e-06,
+      "loss": 3.3081,
+      "step": 2714624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.549466280301606e-06,
+      "loss": 3.3176,
+      "step": 2715136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.541096711594818e-06,
+      "loss": 3.3223,
+      "step": 2715648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.532710764084297e-06,
+      "loss": 3.3144,
+      "step": 2716160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.524324816573778e-06,
+      "loss": 3.3257,
+      "step": 2716672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.5159388690632575e-06,
+      "loss": 3.3171,
+      "step": 2717184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.507569300356469e-06,
+      "loss": 3.3121,
+      "step": 2717696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.499183352845948e-06,
+      "loss": 3.3188,
+      "step": 2718208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.490797405335428e-06,
+      "loss": 3.327,
+      "step": 2718720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.482411457824908e-06,
+      "loss": 3.324,
+      "step": 2719232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.474041889118119e-06,
+      "loss": 3.3331,
+      "step": 2719744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.4656559416076e-06,
+      "loss": 3.3172,
+      "step": 2720256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.457269994097079e-06,
+      "loss": 3.3168,
+      "step": 2720768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.44888404658656e-06,
+      "loss": 3.3195,
+      "step": 2721280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.4405144778797705e-06,
+      "loss": 3.3184,
+      "step": 2721792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.4321285303692505e-06,
+      "loss": 3.3264,
+      "step": 2722304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.42374258285873e-06,
+      "loss": 3.3182,
+      "step": 2722816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.41535663534821e-06,
+      "loss": 3.3169,
+      "step": 2723328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.406987066641422e-06,
+      "loss": 3.3228,
+      "step": 2723840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.398601119130901e-06,
+      "loss": 3.3281,
+      "step": 2724352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.390215171620382e-06,
+      "loss": 3.3112,
+      "step": 2724864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.381829224109861e-06,
+      "loss": 3.3197,
+      "step": 2725376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.373459655403073e-06,
+      "loss": 3.3196,
+      "step": 2725888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.365073707892553e-06,
+      "loss": 3.3135,
+      "step": 2726400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.3566877603820326e-06,
+      "loss": 3.3068,
+      "step": 2726912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.3483018128715125e-06,
+      "loss": 3.3155,
+      "step": 2727424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.339932244164724e-06,
+      "loss": 3.3186,
+      "step": 2727936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.331546296654204e-06,
+      "loss": 3.3172,
+      "step": 2728448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.323160349143683e-06,
+      "loss": 3.3171,
+      "step": 2728960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.314774401633164e-06,
+      "loss": 3.3162,
+      "step": 2729472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.306404832926375e-06,
+      "loss": 3.3369,
+      "step": 2729984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.298018885415855e-06,
+      "loss": 3.3307,
+      "step": 2730496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.289632937905335e-06,
+      "loss": 3.3247,
+      "step": 2731008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.281246990394815e-06,
+      "loss": 3.322,
+      "step": 2731520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.272893800491757e-06,
+      "loss": 3.3169,
+      "step": 2732032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.264507852981237e-06,
+      "loss": 3.3133,
+      "step": 2732544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.256121905470717e-06,
+      "loss": 3.313,
+      "step": 2733056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.247735957960197e-06,
+      "loss": 3.3127,
+      "step": 2733568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.239350010449677e-06,
+      "loss": 3.3203,
+      "step": 2734080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.230964062939157e-06,
+      "loss": 3.3269,
+      "step": 2734592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.222578115428637e-06,
+      "loss": 3.3183,
+      "step": 2735104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.214192167918117e-06,
+      "loss": 3.3248,
+      "step": 2735616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.2058225992113285e-06,
+      "loss": 3.3157,
+      "step": 2736128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.197436651700808e-06,
+      "loss": 3.316,
+      "step": 2736640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.189050704190288e-06,
+      "loss": 3.312,
+      "step": 2737152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.1806647566797675e-06,
+      "loss": 3.3303,
+      "step": 2737664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.172295187972979e-06,
+      "loss": 3.337,
+      "step": 2738176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.163909240462459e-06,
+      "loss": 3.3241,
+      "step": 2738688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.155523292951939e-06,
+      "loss": 3.3184,
+      "step": 2739200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.147137345441419e-06,
+      "loss": 3.3225,
+      "step": 2739712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.13876777673463e-06,
+      "loss": 3.319,
+      "step": 2740224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.130381829224111e-06,
+      "loss": 3.3296,
+      "step": 2740736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.12199588171359e-06,
+      "loss": 3.3163,
+      "step": 2741248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.1136099342030705e-06,
+      "loss": 3.3258,
+      "step": 2741760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.105240365496281e-06,
+      "loss": 3.3172,
+      "step": 2742272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.096854417985761e-06,
+      "loss": 3.3251,
+      "step": 2742784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.088468470475241e-06,
+      "loss": 3.3175,
+      "step": 2743296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.08008252296472e-06,
+      "loss": 3.3196,
+      "step": 2743808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.071712954257932e-06,
+      "loss": 3.3263,
+      "step": 2744320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.063327006747412e-06,
+      "loss": 3.3147,
+      "step": 2744832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.054941059236893e-06,
+      "loss": 3.3233,
+      "step": 2745344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.046555111726372e-06,
+      "loss": 3.3304,
+      "step": 2745856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.0381855430195835e-06,
+      "loss": 3.3223,
+      "step": 2746368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.0297995955090634e-06,
+      "loss": 3.3179,
+      "step": 2746880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.021413647998543e-06,
+      "loss": 3.316,
+      "step": 2747392
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8615188598632812,
+      "eval_runtime": 309.3061,
+      "eval_samples_per_second": 1233.7,
+      "eval_steps_per_second": 38.554,
+      "step": 2747520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.013027700488023e-06,
+      "loss": 3.3171,
+      "step": 2747904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.004658131781235e-06,
+      "loss": 3.3146,
+      "step": 2748416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.996288563074446e-06,
+      "loss": 3.3284,
+      "step": 2748928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.987902615563926e-06,
+      "loss": 3.324,
+      "step": 2749440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.979516668053406e-06,
+      "loss": 3.3347,
+      "step": 2749952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.971130720542886e-06,
+      "loss": 3.3135,
+      "step": 2750464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.962744773032366e-06,
+      "loss": 3.3228,
+      "step": 2750976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.9543588255218455e-06,
+      "loss": 3.3024,
+      "step": 2751488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.945972878011325e-06,
+      "loss": 3.3256,
+      "step": 2752000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.9375869305008055e-06,
+      "loss": 3.3111,
+      "step": 2752512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.929217361794016e-06,
+      "loss": 3.33,
+      "step": 2753024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.920831414283496e-06,
+      "loss": 3.3265,
+      "step": 2753536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.912445466772976e-06,
+      "loss": 3.3039,
+      "step": 2754048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.904059519262456e-06,
+      "loss": 3.3145,
+      "step": 2754560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.895689950555668e-06,
+      "loss": 3.3173,
+      "step": 2755072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.887304003045148e-06,
+      "loss": 3.3068,
+      "step": 2755584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.878918055534628e-06,
+      "loss": 3.3149,
+      "step": 2756096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870532108024107e-06,
+      "loss": 3.3172,
+      "step": 2756608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862162539317318e-06,
+      "loss": 3.3166,
+      "step": 2757120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85379297061053e-06,
+      "loss": 3.3341,
+      "step": 2757632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.84540702310001e-06,
+      "loss": 3.317,
+      "step": 2758144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.837021075589489e-06,
+      "loss": 3.3193,
+      "step": 2758656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.82863512807897e-06,
+      "loss": 3.3211,
+      "step": 2759168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.820249180568449e-06,
+      "loss": 3.3228,
+      "step": 2759680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.81186323305793e-06,
+      "loss": 3.3118,
+      "step": 2760192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.803477285547409e-06,
+      "loss": 3.3229,
+      "step": 2760704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.795091338036889e-06,
+      "loss": 3.3108,
+      "step": 2761216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7867217693301005e-06,
+      "loss": 3.3143,
+      "step": 2761728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7783358218195805e-06,
+      "loss": 3.3098,
+      "step": 2762240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7699498743090604e-06,
+      "loss": 3.3177,
+      "step": 2762752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.76156392679854e-06,
+      "loss": 3.3179,
+      "step": 2763264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.753194358091752e-06,
+      "loss": 3.3212,
+      "step": 2763776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.744808410581231e-06,
+      "loss": 3.3185,
+      "step": 2764288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.736422463070712e-06,
+      "loss": 3.3281,
+      "step": 2764800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.728036515560191e-06,
+      "loss": 3.3111,
+      "step": 2765312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.719666946853403e-06,
+      "loss": 3.3205,
+      "step": 2765824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.711280999342883e-06,
+      "loss": 3.3112,
+      "step": 2766336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.702895051832363e-06,
+      "loss": 3.3161,
+      "step": 2766848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6945091043218426e-06,
+      "loss": 3.3074,
+      "step": 2767360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.686139535615053e-06,
+      "loss": 3.3097,
+      "step": 2767872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.677753588104534e-06,
+      "loss": 3.3123,
+      "step": 2768384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.669367640594013e-06,
+      "loss": 3.3265,
+      "step": 2768896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.660981693083494e-06,
+      "loss": 3.3087,
+      "step": 2769408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.652612124376705e-06,
+      "loss": 3.3135,
+      "step": 2769920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.644226176866185e-06,
+      "loss": 3.3305,
+      "step": 2770432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.635840229355665e-06,
+      "loss": 3.3222,
+      "step": 2770944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.627454281845144e-06,
+      "loss": 3.3191,
+      "step": 2771456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6190847131383555e-06,
+      "loss": 3.3077,
+      "step": 2771968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6106987656278355e-06,
+      "loss": 3.299,
+      "step": 2772480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.602312818117316e-06,
+      "loss": 3.3223,
+      "step": 2772992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593926870606795e-06,
+      "loss": 3.3175,
+      "step": 2773504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585557301900007e-06,
+      "loss": 3.3072,
+      "step": 2774016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577171354389487e-06,
+      "loss": 3.3073,
+      "step": 2774528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568801785682699e-06,
+      "loss": 3.3176,
+      "step": 2775040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560415838172178e-06,
+      "loss": 3.2982,
+      "step": 2775552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5520298906616585e-06,
+      "loss": 3.3175,
+      "step": 2776064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.543643943151138e-06,
+      "loss": 3.3024,
+      "step": 2776576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.535257995640618e-06,
+      "loss": 3.3141,
+      "step": 2777088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5268720481300975e-06,
+      "loss": 3.3158,
+      "step": 2777600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5184861006195775e-06,
+      "loss": 3.3075,
+      "step": 2778112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5101001531090574e-06,
+      "loss": 3.3136,
+      "step": 2778624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.501730584402269e-06,
+      "loss": 3.3116,
+      "step": 2779136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.493344636891749e-06,
+      "loss": 3.3044,
+      "step": 2779648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.484958689381229e-06,
+      "loss": 3.3087,
+      "step": 2780160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.476572741870708e-06,
+      "loss": 3.3095,
+      "step": 2780672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.46820317316392e-06,
+      "loss": 3.3095,
+      "step": 2781184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4598172256534e-06,
+      "loss": 3.3109,
+      "step": 2781696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45143127814288e-06,
+      "loss": 3.3012,
+      "step": 2782208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44304533063236e-06,
+      "loss": 3.3039,
+      "step": 2782720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434675761925571e-06,
+      "loss": 3.3142,
+      "step": 2783232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.426289814415051e-06,
+      "loss": 3.3245,
+      "step": 2783744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.417903866904531e-06,
+      "loss": 3.3072,
+      "step": 2784256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.409517919394011e-06,
+      "loss": 3.3255,
+      "step": 2784768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.401148350687223e-06,
+      "loss": 3.3186,
+      "step": 2785280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.392762403176702e-06,
+      "loss": 3.3111,
+      "step": 2785792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.384376455666182e-06,
+      "loss": 3.315,
+      "step": 2786304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.375990508155662e-06,
+      "loss": 3.3209,
+      "step": 2786816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3676209394488726e-06,
+      "loss": 3.3129,
+      "step": 2787328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.359251370742084e-06,
+      "loss": 3.3125,
+      "step": 2787840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.350865423231564e-06,
+      "loss": 3.3204,
+      "step": 2788352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342479475721044e-06,
+      "loss": 3.3158,
+      "step": 2788864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334093528210524e-06,
+      "loss": 3.3166,
+      "step": 2789376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325707580700004e-06,
+      "loss": 3.3105,
+      "step": 2789888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317321633189484e-06,
+      "loss": 3.2977,
+      "step": 2790400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308935685678964e-06,
+      "loss": 3.2961,
+      "step": 2790912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300549738168444e-06,
+      "loss": 3.3103,
+      "step": 2791424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.292180169461655e-06,
+      "loss": 3.3135,
+      "step": 2791936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2837942219511355e-06,
+      "loss": 3.3081,
+      "step": 2792448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.275408274440615e-06,
+      "loss": 3.3143,
+      "step": 2792960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2670223269300945e-06,
+      "loss": 3.3077,
+      "step": 2793472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258652758223306e-06,
+      "loss": 3.3053,
+      "step": 2793984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250266810712786e-06,
+      "loss": 3.3089,
+      "step": 2794496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.241880863202266e-06,
+      "loss": 3.3121,
+      "step": 2795008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.233494915691746e-06,
+      "loss": 3.3201,
+      "step": 2795520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.225125346984958e-06,
+      "loss": 3.3184,
+      "step": 2796032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.216739399474437e-06,
+      "loss": 3.3153,
+      "step": 2796544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.208353451963918e-06,
+      "loss": 3.3049,
+      "step": 2797056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.199967504453397e-06,
+      "loss": 3.317,
+      "step": 2797568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.191597935746608e-06,
+      "loss": 3.3098,
+      "step": 2798080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.183211988236088e-06,
+      "loss": 3.3181,
+      "step": 2798592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174826040725567e-06,
+      "loss": 3.3138,
+      "step": 2799104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166440093215048e-06,
+      "loss": 3.3046,
+      "step": 2799616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158070524508259e-06,
+      "loss": 3.3192,
+      "step": 2800128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149684576997739e-06,
+      "loss": 3.3163,
+      "step": 2800640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141298629487219e-06,
+      "loss": 3.3082,
+      "step": 2801152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1329126819767e-06,
+      "loss": 3.3145,
+      "step": 2801664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1245431132699105e-06,
+      "loss": 3.3132,
+      "step": 2802176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1161571657593905e-06,
+      "loss": 3.2972,
+      "step": 2802688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.10777121824887e-06,
+      "loss": 3.3004,
+      "step": 2803200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0993852707383495e-06,
+      "loss": 3.3117,
+      "step": 2803712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.091015702031561e-06,
+      "loss": 3.3093,
+      "step": 2804224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.082629754521041e-06,
+      "loss": 3.3093,
+      "step": 2804736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.074243807010521e-06,
+      "loss": 3.3089,
+      "step": 2805248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.065857859500001e-06,
+      "loss": 3.3053,
+      "step": 2805760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.057488290793213e-06,
+      "loss": 3.3288,
+      "step": 2806272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049102343282693e-06,
+      "loss": 3.3256,
+      "step": 2806784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040716395772173e-06,
+      "loss": 3.3176,
+      "step": 2807296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0323304482616525e-06,
+      "loss": 3.3137,
+      "step": 2807808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023977258358595e-06,
+      "loss": 3.3093,
+      "step": 2808320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015591310848075e-06,
+      "loss": 3.3024,
+      "step": 2808832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007205363337555e-06,
+      "loss": 3.307,
+      "step": 2809344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.998819415827035e-06,
+      "loss": 3.3081,
+      "step": 2809856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.990433468316514e-06,
+      "loss": 3.3112,
+      "step": 2810368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.982047520805995e-06,
+      "loss": 3.3148,
+      "step": 2810880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.973661573295475e-06,
+      "loss": 3.3112,
+      "step": 2811392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.965275625784955e-06,
+      "loss": 3.3141,
+      "step": 2811904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9569060570781655e-06,
+      "loss": 3.3091,
+      "step": 2812416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.948520109567646e-06,
+      "loss": 3.3075,
+      "step": 2812928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.940134162057125e-06,
+      "loss": 3.3063,
+      "step": 2813440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.931748214546605e-06,
+      "loss": 3.315,
+      "step": 2813952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923378645839817e-06,
+      "loss": 3.3289,
+      "step": 2814464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.914992698329296e-06,
+      "loss": 3.3162,
+      "step": 2814976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906606750818777e-06,
+      "loss": 3.3078,
+      "step": 2815488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898220803308256e-06,
+      "loss": 3.3148,
+      "step": 2816000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889851234601468e-06,
+      "loss": 3.3089,
+      "step": 2816512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881465287090948e-06,
+      "loss": 3.3199,
+      "step": 2817024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8730793395804276e-06,
+      "loss": 3.3085,
+      "step": 2817536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8646933920699075e-06,
+      "loss": 3.3135,
+      "step": 2818048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.856323823363119e-06,
+      "loss": 3.3097,
+      "step": 2818560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.847937875852599e-06,
+      "loss": 3.3222,
+      "step": 2819072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.839551928342078e-06,
+      "loss": 3.3065,
+      "step": 2819584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.831165980831559e-06,
+      "loss": 3.3123,
+      "step": 2820096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.82279641212477e-06,
+      "loss": 3.3162,
+      "step": 2820608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.81441046461425e-06,
+      "loss": 3.309,
+      "step": 2821120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8060245171037297e-06,
+      "loss": 3.3145,
+      "step": 2821632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79763856959321e-06,
+      "loss": 3.3211,
+      "step": 2822144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789269000886421e-06,
+      "loss": 3.3164,
+      "step": 2822656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7808830533759013e-06,
+      "loss": 3.3069,
+      "step": 2823168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772497105865381e-06,
+      "loss": 3.308,
+      "step": 2823680
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8601794242858887,
+      "eval_runtime": 307.8833,
+      "eval_samples_per_second": 1239.401,
+      "eval_steps_per_second": 38.732,
+      "step": 2823840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7641111583548603e-06,
+      "loss": 3.3074,
+      "step": 2824192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7557252108443407e-06,
+      "loss": 3.3049,
+      "step": 2824704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747355642137552e-06,
+      "loss": 3.3217,
+      "step": 2825216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738969694627032e-06,
+      "loss": 3.3163,
+      "step": 2825728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730583747116512e-06,
+      "loss": 3.3234,
+      "step": 2826240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7221977996059922e-06,
+      "loss": 3.3098,
+      "step": 2826752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7138118520954718e-06,
+      "loss": 3.3101,
+      "step": 2827264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7054259045849513e-06,
+      "loss": 3.2999,
+      "step": 2827776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6970399570744317e-06,
+      "loss": 3.3097,
+      "step": 2828288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.688654009563911e-06,
+      "loss": 3.3083,
+      "step": 2828800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.680284440857123e-06,
+      "loss": 3.3174,
+      "step": 2829312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6718984933466024e-06,
+      "loss": 3.3179,
+      "step": 2829824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.663512545836082e-06,
+      "loss": 3.2971,
+      "step": 2830336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6551265983255623e-06,
+      "loss": 3.3066,
+      "step": 2830848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6467570296187735e-06,
+      "loss": 3.3109,
+      "step": 2831360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.638371082108254e-06,
+      "loss": 3.2973,
+      "step": 2831872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6299851345977334e-06,
+      "loss": 3.3074,
+      "step": 2832384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6215991870872138e-06,
+      "loss": 3.3077,
+      "step": 2832896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6132296183804246e-06,
+      "loss": 3.3098,
+      "step": 2833408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604843670869905e-06,
+      "loss": 3.3231,
+      "step": 2833920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5964577233593845e-06,
+      "loss": 3.3107,
+      "step": 2834432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.588071775848864e-06,
+      "loss": 3.3153,
+      "step": 2834944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5797022071420757e-06,
+      "loss": 3.3104,
+      "step": 2835456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.571316259631555e-06,
+      "loss": 3.3139,
+      "step": 2835968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5629303121210356e-06,
+      "loss": 3.3024,
+      "step": 2836480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5545443646105155e-06,
+      "loss": 3.3141,
+      "step": 2836992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.546174795903727e-06,
+      "loss": 3.3093,
+      "step": 2837504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5377888483932067e-06,
+      "loss": 3.3025,
+      "step": 2838016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.529402900882687e-06,
+      "loss": 3.3019,
+      "step": 2838528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5210169533721666e-06,
+      "loss": 3.3102,
+      "step": 2839040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5126473846653782e-06,
+      "loss": 3.3114,
+      "step": 2839552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5042614371548578e-06,
+      "loss": 3.3117,
+      "step": 2840064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4958754896443373e-06,
+      "loss": 3.3131,
+      "step": 2840576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4874895421338177e-06,
+      "loss": 3.3185,
+      "step": 2841088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.479119973427029e-06,
+      "loss": 3.3043,
+      "step": 2841600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470734025916509e-06,
+      "loss": 3.3099,
+      "step": 2842112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462348078405989e-06,
+      "loss": 3.3064,
+      "step": 2842624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4539621308954688e-06,
+      "loss": 3.3077,
+      "step": 2843136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44559256218868e-06,
+      "loss": 3.2968,
+      "step": 2843648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4372066146781604e-06,
+      "loss": 3.2983,
+      "step": 2844160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42882066716764e-06,
+      "loss": 3.3072,
+      "step": 2844672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4204510984608515e-06,
+      "loss": 3.3165,
+      "step": 2845184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.412065150950331e-06,
+      "loss": 3.303,
+      "step": 2845696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4036792034398106e-06,
+      "loss": 3.3036,
+      "step": 2846208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.395293255929291e-06,
+      "loss": 3.3226,
+      "step": 2846720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.386923687222502e-06,
+      "loss": 3.3129,
+      "step": 2847232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.378537739711982e-06,
+      "loss": 3.3124,
+      "step": 2847744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.370151792201462e-06,
+      "loss": 3.3017,
+      "step": 2848256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.361765844690942e-06,
+      "loss": 3.2902,
+      "step": 2848768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3533962759841533e-06,
+      "loss": 3.3143,
+      "step": 2849280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3450103284736337e-06,
+      "loss": 3.3093,
+      "step": 2849792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336624380963113e-06,
+      "loss": 3.2985,
+      "step": 2850304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3282384334525927e-06,
+      "loss": 3.2997,
+      "step": 2850816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3198688647458044e-06,
+      "loss": 3.3118,
+      "step": 2851328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.311482917235284e-06,
+      "loss": 3.285,
+      "step": 2851840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3030969697247643e-06,
+      "loss": 3.3088,
+      "step": 2852352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.294711022214244e-06,
+      "loss": 3.299,
+      "step": 2852864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.286325074703724e-06,
+      "loss": 3.3023,
+      "step": 2853376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2779391271932037e-06,
+      "loss": 3.3087,
+      "step": 2853888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2695531796826837e-06,
+      "loss": 3.3001,
+      "step": 2854400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2611672321721636e-06,
+      "loss": 3.301,
+      "step": 2854912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.252797663465375e-06,
+      "loss": 3.3065,
+      "step": 2855424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2444117159548552e-06,
+      "loss": 3.2982,
+      "step": 2855936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2360257684443347e-06,
+      "loss": 3.2957,
+      "step": 2856448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.227639820933815e-06,
+      "loss": 3.3031,
+      "step": 2856960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219270252227026e-06,
+      "loss": 3.3049,
+      "step": 2857472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2108843047165063e-06,
+      "loss": 3.2985,
+      "step": 2857984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202498357205986e-06,
+      "loss": 3.2978,
+      "step": 2858496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1941124096954654e-06,
+      "loss": 3.2949,
+      "step": 2859008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185742840988677e-06,
+      "loss": 3.3026,
+      "step": 2859520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177356893478157e-06,
+      "loss": 3.3149,
+      "step": 2860032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1689709459676373e-06,
+      "loss": 3.3012,
+      "step": 2860544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.160584998457117e-06,
+      "loss": 3.317,
+      "step": 2861056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1522154297503285e-06,
+      "loss": 3.3133,
+      "step": 2861568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.143829482239808e-06,
+      "loss": 3.3061,
+      "step": 2862080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1354435347292884e-06,
+      "loss": 3.2993,
+      "step": 2862592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.127057587218768e-06,
+      "loss": 3.3138,
+      "step": 2863104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.118688018511979e-06,
+      "loss": 3.3048,
+      "step": 2863616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.110302071001459e-06,
+      "loss": 3.3014,
+      "step": 2864128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.101916123490939e-06,
+      "loss": 3.3163,
+      "step": 2864640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093530175980419e-06,
+      "loss": 3.3073,
+      "step": 2865152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0851606072736307e-06,
+      "loss": 3.3098,
+      "step": 2865664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07677465976311e-06,
+      "loss": 3.3041,
+      "step": 2866176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06838871225259e-06,
+      "loss": 3.2908,
+      "step": 2866688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0600191435458014e-06,
+      "loss": 3.2856,
+      "step": 2867200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0516331960352813e-06,
+      "loss": 3.3026,
+      "step": 2867712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0432472485247613e-06,
+      "loss": 3.3052,
+      "step": 2868224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0348613010142412e-06,
+      "loss": 3.3,
+      "step": 2868736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0264917323074524e-06,
+      "loss": 3.3046,
+      "step": 2869248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0181057847969324e-06,
+      "loss": 3.3034,
+      "step": 2869760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0097198372864124e-06,
+      "loss": 3.2952,
+      "step": 2870272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0013338897758923e-06,
+      "loss": 3.3045,
+      "step": 2870784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.992964321069104e-06,
+      "loss": 3.3006,
+      "step": 2871296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9845783735585835e-06,
+      "loss": 3.3127,
+      "step": 2871808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9761924260480634e-06,
+      "loss": 3.313,
+      "step": 2872320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9678064785375434e-06,
+      "loss": 3.3071,
+      "step": 2872832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9594369098307546e-06,
+      "loss": 3.2984,
+      "step": 2873344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9510509623202346e-06,
+      "loss": 3.3046,
+      "step": 2873856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9426650148097145e-06,
+      "loss": 3.3018,
+      "step": 2874368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9342790672991945e-06,
+      "loss": 3.3134,
+      "step": 2874880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9259094985924057e-06,
+      "loss": 3.3031,
+      "step": 2875392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9175235510818856e-06,
+      "loss": 3.2989,
+      "step": 2875904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9091376035713656e-06,
+      "loss": 3.3088,
+      "step": 2876416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9007516560608456e-06,
+      "loss": 3.3066,
+      "step": 2876928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8923820873540568e-06,
+      "loss": 3.3011,
+      "step": 2877440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8839961398435367e-06,
+      "loss": 3.3064,
+      "step": 2877952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8756101923330167e-06,
+      "loss": 3.3011,
+      "step": 2878464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8672242448224966e-06,
+      "loss": 3.2985,
+      "step": 2878976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.858854676115708e-06,
+      "loss": 3.2878,
+      "step": 2879488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.850468728605188e-06,
+      "loss": 3.3052,
+      "step": 2880000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8420827810946673e-06,
+      "loss": 3.3062,
+      "step": 2880512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8336968335841473e-06,
+      "loss": 3.3008,
+      "step": 2881024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.825327264877359e-06,
+      "loss": 3.3018,
+      "step": 2881536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.816941317366839e-06,
+      "loss": 3.2964,
+      "step": 2882048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.808555369856319e-06,
+      "loss": 3.3191,
+      "step": 2882560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.800169422345799e-06,
+      "loss": 3.3177,
+      "step": 2883072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.79179985363901e-06,
+      "loss": 3.3102,
+      "step": 2883584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.78341390612849e-06,
+      "loss": 3.307,
+      "step": 2884096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.775044337421701e-06,
+      "loss": 3.2994,
+      "step": 2884608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.766658389911181e-06,
+      "loss": 3.2965,
+      "step": 2885120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.758272442400661e-06,
+      "loss": 3.2977,
+      "step": 2885632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7498864948901406e-06,
+      "loss": 3.2987,
+      "step": 2886144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7415005473796206e-06,
+      "loss": 3.309,
+      "step": 2886656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7331145998691005e-06,
+      "loss": 3.305,
+      "step": 2887168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7247286523585805e-06,
+      "loss": 3.3065,
+      "step": 2887680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.716342704848061e-06,
+      "loss": 3.3019,
+      "step": 2888192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.707973136141272e-06,
+      "loss": 3.3047,
+      "step": 2888704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.699587188630752e-06,
+      "loss": 3.299,
+      "step": 2889216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6912012411202316e-06,
+      "loss": 3.3012,
+      "step": 2889728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6828152936097115e-06,
+      "loss": 3.3021,
+      "step": 2890240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6744457249029227e-06,
+      "loss": 3.3192,
+      "step": 2890752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6660597773924027e-06,
+      "loss": 3.3145,
+      "step": 2891264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6576738298818827e-06,
+      "loss": 3.2988,
+      "step": 2891776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6492878823713626e-06,
+      "loss": 3.3077,
+      "step": 2892288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6409183136645742e-06,
+      "loss": 3.3055,
+      "step": 2892800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.632532366154054e-06,
+      "loss": 3.3087,
+      "step": 2893312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.624146418643534e-06,
+      "loss": 3.3012,
+      "step": 2893824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6157604711330137e-06,
+      "loss": 3.3067,
+      "step": 2894336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6073909024262253e-06,
+      "loss": 3.3011,
+      "step": 2894848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.599004954915705e-06,
+      "loss": 3.3102,
+      "step": 2895360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.590619007405185e-06,
+      "loss": 3.2998,
+      "step": 2895872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5822330598946648e-06,
+      "loss": 3.3071,
+      "step": 2896384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.573863491187876e-06,
+      "loss": 3.3056,
+      "step": 2896896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.565477543677356e-06,
+      "loss": 3.305,
+      "step": 2897408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.557091596166836e-06,
+      "loss": 3.306,
+      "step": 2897920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.548705648656316e-06,
+      "loss": 3.3075,
+      "step": 2898432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5403360799495275e-06,
+      "loss": 3.3179,
+      "step": 2898944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.531950132439007e-06,
+      "loss": 3.2955,
+      "step": 2899456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.523564184928487e-06,
+      "loss": 3.3024,
+      "step": 2899968
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.860586166381836,
+      "eval_runtime": 302.5898,
+      "eval_samples_per_second": 1261.083,
+      "eval_steps_per_second": 39.41,
+      "step": 2900160
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.018438125493078e+18,
+  "trial_name": null,
+  "trial_params": null
+}