diff --git "a/checkpoint-1450080/trainer_state.json" "b/checkpoint-1450080/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1450080/trainer_state.json"
@@ -0,0 +1,17169 @@
+{
+  "best_metric": 3.8445136547088623,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-case/transformer/1/checkpoints/checkpoint-915840",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1450080,
+  "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": 10.8546,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8152,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1962,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9782,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8176,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7102,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6077,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5352,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4531,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4052,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3446,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3026,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2566,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2091,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1715,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1284,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1004,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0707,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0435,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0123,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 4.9999,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.9718,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9432,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9305,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9198,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.8831,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8646,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8512,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.837,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8149,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7976,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.795,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.779,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7723,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7578,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7399,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7336,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.6992,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7062,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6791,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6786,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6663,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.669,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6467,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6443,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6357,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6189,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6294,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5942,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.586,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.5986,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5931,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5698,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5621,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5636,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5446,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5664,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5268,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5372,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5277,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5117,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5193,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5016,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4985,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4896,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5032,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4757,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4807,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4591,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.467,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4785,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4677,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4509,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4635,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.465,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4384,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4448,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4272,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4182,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4163,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.416,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4229,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4289,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930422841748654e-05,
+      "loss": 4.4166,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3931,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.4001,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4018,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927070100624819e-05,
+      "loss": 4.3927,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3842,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3814,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3753,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.383,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9228804026303046e-05,
+      "loss": 4.3712,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922041807879253e-05,
+      "loss": 4.3769,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921203213128201e-05,
+      "loss": 4.3674,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3648,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3516,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3519,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3505,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3549,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3532,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3338,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3418,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3331,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3355,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3227,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3201,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.327,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3172,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.314,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.307,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3036,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9061134212503846e-05,
+      "loss": 4.3163,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.3036,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3029,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.307,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.313,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.2977,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2856,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.2913,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2988,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.2832,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2806,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2808,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2822,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2753,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2792,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2714,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2714,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2639,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.282,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2663,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2584,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2544,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.264,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2577,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885995336627002e-05,
+      "loss": 4.2639,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2487,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2566,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2496,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2508,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2409,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2506,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.237,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879289854379332e-05,
+      "loss": 4.2273,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87845125962828e-05,
+      "loss": 4.2476,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.2333,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.2325,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875935475375124e-05,
+      "loss": 4.2383,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.2219,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.19382905960083,
+      "eval_runtime": 350.9337,
+      "eval_samples_per_second": 1087.359,
+      "eval_steps_per_second": 33.981,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2137,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2158,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2212,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2153,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.2214,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870068587878506e-05,
+      "loss": 4.2092,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869229993127454e-05,
+      "loss": 4.2096,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.2015,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2152,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.1981,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658772520036196e-05,
+      "loss": 4.1967,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650386572525676e-05,
+      "loss": 4.1991,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642017003818885e-05,
+      "loss": 4.1946,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633631056308365e-05,
+      "loss": 4.1878,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625245108797845e-05,
+      "loss": 4.1881,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616859161287324e-05,
+      "loss": 4.1725,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608473213776804e-05,
+      "loss": 4.1837,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.1761,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.1701,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.1965,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857494580253846e-05,
+      "loss": 4.1733,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856655985502793e-05,
+      "loss": 4.1776,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855817390751741e-05,
+      "loss": 4.1751,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854978796000689e-05,
+      "loss": 4.1891,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854141839130011e-05,
+      "loss": 4.1604,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853303244378958e-05,
+      "loss": 4.1647,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852464649627906e-05,
+      "loss": 4.1694,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851626054876854e-05,
+      "loss": 4.1674,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850789098006175e-05,
+      "loss": 4.1577,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849950503255123e-05,
+      "loss": 4.1559,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849111908504072e-05,
+      "loss": 4.1627,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84827331375302e-05,
+      "loss": 4.1628,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847434719001968e-05,
+      "loss": 4.1662,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846597762131289e-05,
+      "loss": 4.1563,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845759167380237e-05,
+      "loss": 4.1519,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844920572629185e-05,
+      "loss": 4.1597,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844081977878133e-05,
+      "loss": 4.1329,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432450210074536e-05,
+      "loss": 4.1532,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8424064262564016e-05,
+      "loss": 4.1302,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415678315053496e-05,
+      "loss": 4.1429,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8407292367542976e-05,
+      "loss": 4.1369,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398922798836185e-05,
+      "loss": 4.1466,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839053685132567e-05,
+      "loss": 4.1406,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838215090381515e-05,
+      "loss": 4.1317,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837376495630463e-05,
+      "loss": 4.141,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836539538759784e-05,
+      "loss": 4.1318,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835700944008732e-05,
+      "loss": 4.142,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83486234925768e-05,
+      "loss": 4.117,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834025392387001e-05,
+      "loss": 4.1176,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833186797635949e-05,
+      "loss": 4.1338,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832348202884897e-05,
+      "loss": 4.1343,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831509608133845e-05,
+      "loss": 4.1175,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830671013382793e-05,
+      "loss": 4.1149,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829832418631741e-05,
+      "loss": 4.1232,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828993823880689e-05,
+      "loss": 4.1059,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828155229129637e-05,
+      "loss": 4.1311,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273182722589585e-05,
+      "loss": 4.1108,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264796775079065e-05,
+      "loss": 4.1188,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256410827568545e-05,
+      "loss": 4.1123,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248024880058025e-05,
+      "loss": 4.1021,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239655311351234e-05,
+      "loss": 4.116,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231269363840714e-05,
+      "loss": 4.1023,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222883416330194e-05,
+      "loss": 4.101,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214497468819674e-05,
+      "loss": 4.0972,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820612790011288e-05,
+      "loss": 4.1172,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819774195260236e-05,
+      "loss": 4.0964,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818935600509184e-05,
+      "loss": 4.0988,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818097005758132e-05,
+      "loss": 4.0884,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817260048887454e-05,
+      "loss": 4.0962,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816421454136402e-05,
+      "loss": 4.1103,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81558285938535e-05,
+      "loss": 4.1056,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814744264634298e-05,
+      "loss": 4.0938,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813907307763619e-05,
+      "loss": 4.1045,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813068713012567e-05,
+      "loss": 4.1081,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812230118261515e-05,
+      "loss": 4.0936,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811391523510463e-05,
+      "loss": 4.0982,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810554566639784e-05,
+      "loss": 4.0889,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809715971888732e-05,
+      "loss": 4.0787,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887737713768e-05,
+      "loss": 4.0819,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808038782386628e-05,
+      "loss": 4.0854,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807201825515949e-05,
+      "loss": 4.09,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806363230764897e-05,
+      "loss": 4.1036,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805524636013845e-05,
+      "loss": 4.0916,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804686041262793e-05,
+      "loss": 4.0693,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803849084392114e-05,
+      "loss": 4.0814,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803010489641062e-05,
+      "loss": 4.0845,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80217189489001e-05,
+      "loss": 4.0808,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801333300138958e-05,
+      "loss": 4.0733,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800496343268279e-05,
+      "loss": 4.0742,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799657748517227e-05,
+      "loss": 4.0728,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798819153766175e-05,
+      "loss": 4.0756,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797980559015123e-05,
+      "loss": 4.07,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797141964264071e-05,
+      "loss": 4.0765,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796303369513019e-05,
+      "loss": 4.0719,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954664126423406e-05,
+      "loss": 4.0758,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946278178912886e-05,
+      "loss": 4.0591,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793789223140236e-05,
+      "loss": 4.0698,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792950628389184e-05,
+      "loss": 4.0661,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792112033638132e-05,
+      "loss": 4.0693,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79127343888708e-05,
+      "loss": 4.0709,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790434844136028e-05,
+      "loss": 4.0593,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789597887265349e-05,
+      "loss": 4.0607,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788759292514297e-05,
+      "loss": 4.0603,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787920697763245e-05,
+      "loss": 4.0641,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870821030121935e-05,
+      "loss": 4.0493,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862451461415144e-05,
+      "loss": 4.0517,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854065513904624e-05,
+      "loss": 4.0618,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845679566394104e-05,
+      "loss": 4.0504,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783730999768731e-05,
+      "loss": 4.0511,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782892405017679e-05,
+      "loss": 4.0465,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782053810266627e-05,
+      "loss": 4.0417,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781215215515575e-05,
+      "loss": 4.0648,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780376620764523e-05,
+      "loss": 4.045,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779538026013471e-05,
+      "loss": 4.0533,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778699431262419e-05,
+      "loss": 4.0499,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777860836511367e-05,
+      "loss": 4.0656,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777023879640689e-05,
+      "loss": 4.051,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776185284889637e-05,
+      "loss": 4.0368,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775346690138585e-05,
+      "loss": 4.0452,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774508095387533e-05,
+      "loss": 4.052,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773671138516854e-05,
+      "loss": 4.036,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728341816461747e-05,
+      "loss": 4.0426,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719955868951227e-05,
+      "loss": 4.0367,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711569921440707e-05,
+      "loss": 4.0478,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703183973930186e-05,
+      "loss": 4.036,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694798026419666e-05,
+      "loss": 4.0447,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686412078909146e-05,
+      "loss": 4.0387,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7678026131398626e-05,
+      "loss": 4.0405,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766965656269184e-05,
+      "loss": 4.0317,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766127061518132e-05,
+      "loss": 4.0538,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528846676708e-05,
+      "loss": 4.0419,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764449872016028e-05,
+      "loss": 4.0269,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763612915145349e-05,
+      "loss": 4.0294,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762774320394297e-05,
+      "loss": 4.0358,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761935725643245e-05,
+      "loss": 4.037,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761097130892193e-05,
+      "loss": 4.0468,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760260174021514e-05,
+      "loss": 4.0271,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759421579270462e-05,
+      "loss": 4.0368,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75858298451941e-05,
+      "loss": 4.0306,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757744389768358e-05,
+      "loss": 4.0367,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569074328976796e-05,
+      "loss": 4.0284,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560688381466276e-05,
+      "loss": 4.0341,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552302433955756e-05,
+      "loss": 4.0237,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543916486445236e-05,
+      "loss": 4.0192,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535546917738445e-05,
+      "loss": 4.0375,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527160970227925e-05,
+      "loss": 4.0256,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518775022717405e-05,
+      "loss": 4.0242,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510389075206885e-05,
+      "loss": 4.027,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502003127696365e-05,
+      "loss": 4.0179,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.026176452636719,
+      "eval_runtime": 347.0521,
+      "eval_samples_per_second": 1099.521,
+      "eval_steps_per_second": 34.361,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493617180185845e-05,
+      "loss": 4.0104,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748523123267532e-05,
+      "loss": 4.0104,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74768452851648e-05,
+      "loss": 4.0234,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746845933765428e-05,
+      "loss": 4.0124,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460073390143764e-05,
+      "loss": 4.027,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451687442633244e-05,
+      "loss": 4.0102,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443301495122724e-05,
+      "loss": 4.0131,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434915547612204e-05,
+      "loss": 4.0062,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.023,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.0077,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409757705080644e-05,
+      "loss": 4.005,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740138813637385e-05,
+      "loss": 4.0056,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 4.0061,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 3.9997,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 4.0025,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 3.9911,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 3.9964,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 3.9906,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734271926140768e-05,
+      "loss": 3.9877,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0202,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 3.9942,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 3.9924,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 4.0001,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730080590265881e-05,
+      "loss": 4.0074,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292436333952016e-05,
+      "loss": 3.9851,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7284050386441496e-05,
+      "loss": 3.9918,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275664438930976e-05,
+      "loss": 3.9908,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267278491420456e-05,
+      "loss": 3.9955,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258892543909936e-05,
+      "loss": 3.981,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250506596399416e-05,
+      "loss": 3.9856,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 3.9876,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 3.994,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722536513267159e-05,
+      "loss": 4.0011,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721697918516107e-05,
+      "loss": 3.9884,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 3.9814,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 3.9918,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719183772143324e-05,
+      "loss": 3.9722,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718345177392272e-05,
+      "loss": 3.9857,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71750658264122e-05,
+      "loss": 3.9676,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716669625770541e-05,
+      "loss": 3.9759,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715831031019489e-05,
+      "loss": 3.9779,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714992436268437e-05,
+      "loss": 3.9877,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714153841517385e-05,
+      "loss": 3.9805,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 3.9737,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9787,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9753,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710799462513177e-05,
+      "loss": 3.9828,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709962505642498e-05,
+      "loss": 3.962,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091255487718194e-05,
+      "loss": 3.9604,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082869540207674e-05,
+      "loss": 3.9801,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074483592697154e-05,
+      "loss": 3.9756,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706609764518663e-05,
+      "loss": 3.9642,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705772807647984e-05,
+      "loss": 3.9645,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704934212896932e-05,
+      "loss": 3.9684,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70409561814588e-05,
+      "loss": 3.9535,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703257023394828e-05,
+      "loss": 3.9767,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.963,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701579833892724e-05,
+      "loss": 3.9681,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700741239141672e-05,
+      "loss": 3.9616,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69990264439062e-05,
+      "loss": 3.9598,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699065687519941e-05,
+      "loss": 3.9616,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698227092768889e-05,
+      "loss": 3.9593,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697388498017837e-05,
+      "loss": 3.9532,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696549903266785e-05,
+      "loss": 3.9515,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695712946396106e-05,
+      "loss": 3.9704,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694874351645054e-05,
+      "loss": 3.9534,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694035756894003e-05,
+      "loss": 3.9543,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693197162142951e-05,
+      "loss": 3.9463,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692358567391899e-05,
+      "loss": 3.9538,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.9699,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906830157701677e-05,
+      "loss": 3.9646,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898444210191157e-05,
+      "loss": 3.9563,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890058262680636e-05,
+      "loss": 3.9637,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881672315170116e-05,
+      "loss": 3.9676,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873286367659596e-05,
+      "loss": 3.951,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864900420149076e-05,
+      "loss": 3.9603,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856514472638556e-05,
+      "loss": 3.9525,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848144903931765e-05,
+      "loss": 3.9434,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683977533522498e-05,
+      "loss": 3.9507,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683138938771446e-05,
+      "loss": 3.9461,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682300344020394e-05,
+      "loss": 3.9553,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681461749269342e-05,
+      "loss": 3.9636,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680624792398663e-05,
+      "loss": 3.9592,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679786197647611e-05,
+      "loss": 3.9332,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678947602896559e-05,
+      "loss": 3.9493,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678109008145507e-05,
+      "loss": 3.9493,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677270413394455e-05,
+      "loss": 3.9496,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676431818643403e-05,
+      "loss": 3.9436,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675593223892351e-05,
+      "loss": 3.9373,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674756267021672e-05,
+      "loss": 3.9492,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67391767227062e-05,
+      "loss": 3.9407,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673079077519568e-05,
+      "loss": 3.9401,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672240482768516e-05,
+      "loss": 3.9466,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671401888017464e-05,
+      "loss": 3.9419,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705649311467855e-05,
+      "loss": 3.9496,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697263363957335e-05,
+      "loss": 3.9349,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688877416446815e-05,
+      "loss": 3.9382,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668049146893629e-05,
+      "loss": 3.9386,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667210552142577e-05,
+      "loss": 3.9465,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663735952718984e-05,
+      "loss": 3.9444,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665535000520846e-05,
+      "loss": 3.9345,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664696405769794e-05,
+      "loss": 3.9358,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663857811018742e-05,
+      "loss": 3.9356,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663020854148063e-05,
+      "loss": 3.9395,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662182259397011e-05,
+      "loss": 3.9231,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661343664645959e-05,
+      "loss": 3.9274,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660505069894907e-05,
+      "loss": 3.943,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659666475143855e-05,
+      "loss": 3.9329,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658829518273176e-05,
+      "loss": 3.9269,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657990923522124e-05,
+      "loss": 3.9252,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657152328771072e-05,
+      "loss": 3.9241,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65631373402002e-05,
+      "loss": 3.9406,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655475139268968e-05,
+      "loss": 3.9297,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654636544517916e-05,
+      "loss": 3.9292,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653799587647237e-05,
+      "loss": 3.9327,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652960992896185e-05,
+      "loss": 3.9469,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652122398145134e-05,
+      "loss": 3.9334,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651283803394082e-05,
+      "loss": 3.9207,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65044520864303e-05,
+      "loss": 3.9294,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649606613891978e-05,
+      "loss": 3.9308,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487696570212986e-05,
+      "loss": 3.9212,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479310622702466e-05,
+      "loss": 3.9237,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470924675191946e-05,
+      "loss": 3.92,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462538727681426e-05,
+      "loss": 3.9329,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454152780170906e-05,
+      "loss": 3.9175,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445783211464115e-05,
+      "loss": 3.9331,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437397263953595e-05,
+      "loss": 3.916,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429011316443075e-05,
+      "loss": 3.9277,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420625368932555e-05,
+      "loss": 3.9156,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641225580022577e-05,
+      "loss": 3.9374,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640386985271525e-05,
+      "loss": 3.9298,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639548390520473e-05,
+      "loss": 3.9207,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638709795769421e-05,
+      "loss": 3.912,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637871201018369e-05,
+      "loss": 3.9251,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63703424414769e-05,
+      "loss": 3.9218,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636195649396638e-05,
+      "loss": 3.9344,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635357054645586e-05,
+      "loss": 3.9146,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634518459894534e-05,
+      "loss": 3.9282,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633679865143482e-05,
+      "loss": 3.9202,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632841270392429e-05,
+      "loss": 3.9311,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632002675641377e-05,
+      "loss": 3.9145,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631165718770699e-05,
+      "loss": 3.923,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303287619000204e-05,
+      "loss": 3.9164,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294901671489684e-05,
+      "loss": 3.9095,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286515723979164e-05,
+      "loss": 3.9293,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278129776468644e-05,
+      "loss": 3.9225,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626974382895812e-05,
+      "loss": 3.9099,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626137426025133e-05,
+      "loss": 3.9197,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625298831274081e-05,
+      "loss": 3.9092,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.952585220336914,
+      "eval_runtime": 319.2814,
+      "eval_samples_per_second": 1195.156,
+      "eval_steps_per_second": 37.349,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624460236523029e-05,
+      "loss": 3.9082,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236216417719766e-05,
+      "loss": 3.903,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227830470209246e-05,
+      "loss": 3.9131,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219444522698726e-05,
+      "loss": 3.9114,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621105857518821e-05,
+      "loss": 3.9246,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620267262767769e-05,
+      "loss": 3.9005,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61943030589709e-05,
+      "loss": 3.9095,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618591711146038e-05,
+      "loss": 3.9012,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617753116394986e-05,
+      "loss": 3.923,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616914521643934e-05,
+      "loss": 3.9022,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616079202653629e-05,
+      "loss": 3.9002,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615240607902577e-05,
+      "loss": 3.9074,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6144036510318976e-05,
+      "loss": 3.9008,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135650562808456e-05,
+      "loss": 3.8988,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127264615297936e-05,
+      "loss": 3.9002,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118878667787416e-05,
+      "loss": 3.8925,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110492720276896e-05,
+      "loss": 3.8921,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6102106772766376e-05,
+      "loss": 3.8927,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093720825255856e-05,
+      "loss": 3.8879,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085334877745336e-05,
+      "loss": 3.9184,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076965309038545e-05,
+      "loss": 3.8942,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068579361528025e-05,
+      "loss": 3.8944,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060193414017505e-05,
+      "loss": 3.8998,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051807466506985e-05,
+      "loss": 3.9041,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043437897800194e-05,
+      "loss": 3.8899,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035051950289674e-05,
+      "loss": 3.8953,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026666002779154e-05,
+      "loss": 3.8925,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018280055268634e-05,
+      "loss": 3.8957,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009894107758114e-05,
+      "loss": 3.886,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.60015081602476e-05,
+      "loss": 3.8898,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599312221273708e-05,
+      "loss": 3.8951,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598473626522656e-05,
+      "loss": 3.8967,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597636669651977e-05,
+      "loss": 3.9039,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596799712781298e-05,
+      "loss": 3.8954,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595961118030246e-05,
+      "loss": 3.8856,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595122523279194e-05,
+      "loss": 3.8959,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594283928528142e-05,
+      "loss": 3.8813,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59344533377709e-05,
+      "loss": 3.8885,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592608376906411e-05,
+      "loss": 3.8748,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591769782155359e-05,
+      "loss": 3.8819,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590931187404307e-05,
+      "loss": 3.8797,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900925926532554e-05,
+      "loss": 3.897,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892539979022034e-05,
+      "loss": 3.8908,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884154031511514e-05,
+      "loss": 3.878,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587578446280472e-05,
+      "loss": 3.8844,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58673985152942e-05,
+      "loss": 3.8832,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585901256778368e-05,
+      "loss": 3.8902,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585062662027316e-05,
+      "loss": 3.8735,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584224067276264e-05,
+      "loss": 3.8655,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583387110405585e-05,
+      "loss": 3.889,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582548515654533e-05,
+      "loss": 3.8846,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581709920903481e-05,
+      "loss": 3.8695,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580871326152429e-05,
+      "loss": 3.8781,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580034369281751e-05,
+      "loss": 3.8814,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579195774530699e-05,
+      "loss": 3.8639,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578357179779647e-05,
+      "loss": 3.8826,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577518585028595e-05,
+      "loss": 3.8729,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576681628157916e-05,
+      "loss": 3.8838,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758430334068637e-05,
+      "loss": 3.8681,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750044386558117e-05,
+      "loss": 3.8788,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741658439047596e-05,
+      "loss": 3.8695,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733288870340806e-05,
+      "loss": 3.8721,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724902922830286e-05,
+      "loss": 3.8626,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716516975319765e-05,
+      "loss": 3.8647,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708131027809245e-05,
+      "loss": 3.8866,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699745080298725e-05,
+      "loss": 3.8644,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569137551159194e-05,
+      "loss": 3.8687,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568298956408142e-05,
+      "loss": 3.8587,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56746036165709e-05,
+      "loss": 3.8677,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666217669060374e-05,
+      "loss": 3.8824,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657831721549854e-05,
+      "loss": 3.8768,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564946215284307e-05,
+      "loss": 3.8726,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564107620533255e-05,
+      "loss": 3.8804,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563269025782202e-05,
+      "loss": 3.8777,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56243043103115e-05,
+      "loss": 3.8689,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561593474160472e-05,
+      "loss": 3.8739,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56075487940942e-05,
+      "loss": 3.8692,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559916284658368e-05,
+      "loss": 3.863,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559077689907316e-05,
+      "loss": 3.8648,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558239095156264e-05,
+      "loss": 3.8651,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557402138285585e-05,
+      "loss": 3.8692,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556563543534533e-05,
+      "loss": 3.8781,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555724948783481e-05,
+      "loss": 3.8752,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554886354032429e-05,
+      "loss": 3.8493,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554047759281377e-05,
+      "loss": 3.8706,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553210802410698e-05,
+      "loss": 3.8654,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552372207659646e-05,
+      "loss": 3.8668,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551533612908594e-05,
+      "loss": 3.8619,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550695018157542e-05,
+      "loss": 3.8545,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54985642340649e-05,
+      "loss": 3.867,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549019466535811e-05,
+      "loss": 3.8575,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548180871784759e-05,
+      "loss": 3.8597,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547342277033707e-05,
+      "loss": 3.8618,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546503682282655e-05,
+      "loss": 3.8607,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545666725411976e-05,
+      "loss": 3.87,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544828130660924e-05,
+      "loss": 3.8558,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543989535909872e-05,
+      "loss": 3.8543,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54315094115882e-05,
+      "loss": 3.8575,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542312346407768e-05,
+      "loss": 3.8707,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541475389537089e-05,
+      "loss": 3.8618,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540636794786037e-05,
+      "loss": 3.8555,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539798200034985e-05,
+      "loss": 3.8544,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538959605283933e-05,
+      "loss": 3.8577,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381226484132546e-05,
+      "loss": 3.8662,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372840536622026e-05,
+      "loss": 3.8483,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364454589111506e-05,
+      "loss": 3.8445,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356068641600986e-05,
+      "loss": 3.8625,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347699072894195e-05,
+      "loss": 3.852,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339329504187404e-05,
+      "loss": 3.8476,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330943556676884e-05,
+      "loss": 3.8466,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322557609166364e-05,
+      "loss": 3.8492,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314171661655844e-05,
+      "loss": 3.8634,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305785714145324e-05,
+      "loss": 3.8473,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297399766634804e-05,
+      "loss": 3.8529,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5289013819124284e-05,
+      "loss": 3.8572,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528062787161377e-05,
+      "loss": 3.8674,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527225830290698e-05,
+      "loss": 3.8615,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526387235539646e-05,
+      "loss": 3.8418,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525548640788594e-05,
+      "loss": 3.8549,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524710046037542e-05,
+      "loss": 3.8508,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523873089166863e-05,
+      "loss": 3.8482,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523034494415811e-05,
+      "loss": 3.8481,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522195899664759e-05,
+      "loss": 3.8414,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521357304913707e-05,
+      "loss": 3.8543,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520518710162655e-05,
+      "loss": 3.8439,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519681753291976e-05,
+      "loss": 3.8536,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518843158540924e-05,
+      "loss": 3.8439,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5180045637898725e-05,
+      "loss": 3.8504,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171659690388204e-05,
+      "loss": 3.8401,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163290121681414e-05,
+      "loss": 3.8669,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154904174170894e-05,
+      "loss": 3.8504,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146518226660373e-05,
+      "loss": 3.8484,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5138132279149853e-05,
+      "loss": 3.8381,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512976271044306e-05,
+      "loss": 3.8523,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512137676293254e-05,
+      "loss": 3.8466,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511299081542202e-05,
+      "loss": 3.8575,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51046048679115e-05,
+      "loss": 3.8423,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509623529920471e-05,
+      "loss": 3.8539,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508784935169419e-05,
+      "loss": 3.8486,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507946340418368e-05,
+      "loss": 3.861,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507107745667316e-05,
+      "loss": 3.837,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506270788796637e-05,
+      "loss": 3.8514,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505432194045585e-05,
+      "loss": 3.8441,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504593599294533e-05,
+      "loss": 3.8411,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50375500454348e-05,
+      "loss": 3.8516,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502916409792428e-05,
+      "loss": 3.8499,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020794529217496e-05,
+      "loss": 3.8407,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012408581706976e-05,
+      "loss": 3.8441,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500402263419645e-05,
+      "loss": 3.8414,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.912287473678589,
+      "eval_runtime": 322.0035,
+      "eval_samples_per_second": 1185.052,
+      "eval_steps_per_second": 37.034,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499563668668593e-05,
+      "loss": 3.836,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987250739175416e-05,
+      "loss": 3.8294,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978864791664896e-05,
+      "loss": 3.8424,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970478844154376e-05,
+      "loss": 3.8418,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962092896643856e-05,
+      "loss": 3.8532,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953706949133336e-05,
+      "loss": 3.8349,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945321001622816e-05,
+      "loss": 3.8351,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936935054112296e-05,
+      "loss": 3.8326,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928565485405505e-05,
+      "loss": 3.8495,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920179537894985e-05,
+      "loss": 3.8365,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911793590384465e-05,
+      "loss": 3.8285,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903407642873945e-05,
+      "loss": 3.8341,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489505445297088e-05,
+      "loss": 3.8298,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488666850546037e-05,
+      "loss": 3.8321,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487828255794985e-05,
+      "loss": 3.8357,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486989661043933e-05,
+      "loss": 3.8196,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486151066292881e-05,
+      "loss": 3.8254,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485312471541829e-05,
+      "loss": 3.8233,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484473876790777e-05,
+      "loss": 3.8188,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483635282039725e-05,
+      "loss": 3.8465,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482796687288673e-05,
+      "loss": 3.8298,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481958092537621e-05,
+      "loss": 3.8263,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481119497786569e-05,
+      "loss": 3.8303,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480280903035516e-05,
+      "loss": 3.8385,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479443946164838e-05,
+      "loss": 3.8209,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478605351413786e-05,
+      "loss": 3.8282,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477766756662734e-05,
+      "loss": 3.8235,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476928161911682e-05,
+      "loss": 3.8242,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47608956716063e-05,
+      "loss": 3.8189,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752526102899514e-05,
+      "loss": 3.8254,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4744140155388994e-05,
+      "loss": 3.8271,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473575420787847e-05,
+      "loss": 3.8286,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472736826036795e-05,
+      "loss": 3.8372,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471898231285743e-05,
+      "loss": 3.8307,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710612744150636e-05,
+      "loss": 3.8178,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702226796640116e-05,
+      "loss": 3.8287,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693840849129596e-05,
+      "loss": 3.8214,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685454901619076e-05,
+      "loss": 3.8206,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677068954108556e-05,
+      "loss": 3.8107,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466869938540177e-05,
+      "loss": 3.8167,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466031343789125e-05,
+      "loss": 3.8176,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465192749038073e-05,
+      "loss": 3.8294,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464354154287021e-05,
+      "loss": 3.8245,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463515559535969e-05,
+      "loss": 3.8155,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46267860266529e-05,
+      "loss": 3.8191,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461840007914238e-05,
+      "loss": 3.8203,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461001413163186e-05,
+      "loss": 3.8252,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460162818412134e-05,
+      "loss": 3.8095,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459325861541455e-05,
+      "loss": 3.7997,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458487266790403e-05,
+      "loss": 3.8235,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457648672039351e-05,
+      "loss": 3.8201,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456810077288299e-05,
+      "loss": 3.8043,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559731204176205e-05,
+      "loss": 3.8137,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551345256665685e-05,
+      "loss": 3.8195,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542959309155165e-05,
+      "loss": 3.7994,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534573361644645e-05,
+      "loss": 3.8189,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526187414134125e-05,
+      "loss": 3.8104,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517817845427334e-05,
+      "loss": 3.8177,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509431897916814e-05,
+      "loss": 3.8069,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501045950406294e-05,
+      "loss": 3.815,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492660002895774e-05,
+      "loss": 3.8078,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448429043418898e-05,
+      "loss": 3.814,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447590448667846e-05,
+      "loss": 3.7984,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446751853916794e-05,
+      "loss": 3.8057,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445913259165742e-05,
+      "loss": 3.8203,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445076302295064e-05,
+      "loss": 3.8052,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444237707544012e-05,
+      "loss": 3.8047,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44339911279296e-05,
+      "loss": 3.798,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442560518041908e-05,
+      "loss": 3.8038,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441721923290856e-05,
+      "loss": 3.8156,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440884966420177e-05,
+      "loss": 3.8171,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440046371669125e-05,
+      "loss": 3.8135,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439207776918073e-05,
+      "loss": 3.8193,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438369182167021e-05,
+      "loss": 3.8191,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437532225296342e-05,
+      "loss": 3.8065,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669363054529e-05,
+      "loss": 3.8122,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435855035794238e-05,
+      "loss": 3.812,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4350164410431864e-05,
+      "loss": 3.8003,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434179484172507e-05,
+      "loss": 3.8034,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433340889421455e-05,
+      "loss": 3.8053,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432502294670403e-05,
+      "loss": 3.8066,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431663699919351e-05,
+      "loss": 3.8207,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430825105168299e-05,
+      "loss": 3.8128,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42998814829762e-05,
+      "loss": 3.7913,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429149553546568e-05,
+      "loss": 3.8062,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428310958795516e-05,
+      "loss": 3.8088,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427472364044464e-05,
+      "loss": 3.808,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426633769293412e-05,
+      "loss": 3.8045,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425796812422733e-05,
+      "loss": 3.793,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424958217671682e-05,
+      "loss": 3.807,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42411962292063e-05,
+      "loss": 3.7992,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423281028169577e-05,
+      "loss": 3.7999,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224440712988986e-05,
+      "loss": 3.8011,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216054765478466e-05,
+      "loss": 3.8063,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207668817967946e-05,
+      "loss": 3.8073,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419928287045742e-05,
+      "loss": 3.8056,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190913301750635e-05,
+      "loss": 3.7896,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182527354240115e-05,
+      "loss": 3.8021,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174141406729595e-05,
+      "loss": 3.8056,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416575545921907e-05,
+      "loss": 3.808,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157369511708555e-05,
+      "loss": 3.7948,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414899994300177e-05,
+      "loss": 3.797,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140613995491244e-05,
+      "loss": 3.7994,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132228047980724e-05,
+      "loss": 3.806,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123842100470204e-05,
+      "loss": 3.7928,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411547253176342e-05,
+      "loss": 3.7915,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410708658425289e-05,
+      "loss": 3.7994,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409870063674237e-05,
+      "loss": 3.7954,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409033106803559e-05,
+      "loss": 3.7897,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408194512052507e-05,
+      "loss": 3.7919,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407355917301454e-05,
+      "loss": 3.7908,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406517322550402e-05,
+      "loss": 3.8044,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405678727799351e-05,
+      "loss": 3.7921,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404840133048299e-05,
+      "loss": 3.7929,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404001538297247e-05,
+      "loss": 3.7998,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403162943546195e-05,
+      "loss": 3.8123,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402325986675516e-05,
+      "loss": 3.8048,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401487391924464e-05,
+      "loss": 3.7868,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400648797173412e-05,
+      "loss": 3.7945,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39981020242236e-05,
+      "loss": 3.7955,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398973245551681e-05,
+      "loss": 3.7915,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981346508006287e-05,
+      "loss": 3.793,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972960560495767e-05,
+      "loss": 3.785,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964574612985247e-05,
+      "loss": 3.7955,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956205044278456e-05,
+      "loss": 3.7929,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394781909676794e-05,
+      "loss": 3.7987,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393943314925742e-05,
+      "loss": 3.786,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39310472017469e-05,
+      "loss": 3.791,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392267763304011e-05,
+      "loss": 3.7913,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391429168552959e-05,
+      "loss": 3.8042,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390590573801907e-05,
+      "loss": 3.791,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389751979050855e-05,
+      "loss": 3.7978,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388915022180176e-05,
+      "loss": 3.7833,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388076427429124e-05,
+      "loss": 3.8009,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387237832678072e-05,
+      "loss": 3.7938,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38639923792702e-05,
+      "loss": 3.7997,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385560643175968e-05,
+      "loss": 3.785,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847236863052896e-05,
+      "loss": 3.8022,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838850915542376e-05,
+      "loss": 3.7886,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830464968031856e-05,
+      "loss": 3.8064,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822079020521336e-05,
+      "loss": 3.7858,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813709451814545e-05,
+      "loss": 3.797,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805323504304025e-05,
+      "loss": 3.7894,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796937556793505e-05,
+      "loss": 3.7831,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788551609282985e-05,
+      "loss": 3.7991,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780182040576194e-05,
+      "loss": 3.7962,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771796093065674e-05,
+      "loss": 3.7882,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763410145555154e-05,
+      "loss": 3.7862,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3755024198044634e-05,
+      "loss": 3.7919,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.88671612739563,
+      "eval_runtime": 304.1142,
+      "eval_samples_per_second": 1254.762,
+      "eval_steps_per_second": 39.212,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374665462933785e-05,
+      "loss": 3.7814,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373826868182733e-05,
+      "loss": 3.7771,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372988273431681e-05,
+      "loss": 3.7889,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372149678680629e-05,
+      "loss": 3.787,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371311083929577e-05,
+      "loss": 3.7979,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370472489178525e-05,
+      "loss": 3.7821,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369633894427473e-05,
+      "loss": 3.78,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36879529967642e-05,
+      "loss": 3.7793,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367958342805742e-05,
+      "loss": 3.7965,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36711974805469e-05,
+      "loss": 3.7805,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366281153303637e-05,
+      "loss": 3.7794,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365442558552585e-05,
+      "loss": 3.7792,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36460723956228e-05,
+      "loss": 3.7806,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363768644811228e-05,
+      "loss": 3.7788,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362930050060176e-05,
+      "loss": 3.7819,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362091455309124e-05,
+      "loss": 3.7689,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361252860558072e-05,
+      "loss": 3.7696,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36041426580702e-05,
+      "loss": 3.7755,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595756710559676e-05,
+      "loss": 3.7658,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3587370763049156e-05,
+      "loss": 3.793,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357900119434237e-05,
+      "loss": 3.7793,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570615246831845e-05,
+      "loss": 3.7766,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562229299321325e-05,
+      "loss": 3.7746,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3553843351810805e-05,
+      "loss": 3.7863,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354547378310402e-05,
+      "loss": 3.7706,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35370878355935e-05,
+      "loss": 3.7801,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352870188808298e-05,
+      "loss": 3.7696,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352031594057246e-05,
+      "loss": 3.7757,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351194637186568e-05,
+      "loss": 3.7702,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350356042435515e-05,
+      "loss": 3.7722,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349517447684463e-05,
+      "loss": 3.7766,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348678852933411e-05,
+      "loss": 3.7781,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347841896062732e-05,
+      "loss": 3.7846,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34700330131168e-05,
+      "loss": 3.7813,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346164706560628e-05,
+      "loss": 3.7671,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345326111809576e-05,
+      "loss": 3.7796,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444891549388975e-05,
+      "loss": 3.7727,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436505601878455e-05,
+      "loss": 3.7629,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3428119654367935e-05,
+      "loss": 3.7672,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419733706857415e-05,
+      "loss": 3.7645,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411364138150624e-05,
+      "loss": 3.7679,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402978190640104e-05,
+      "loss": 3.7818,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394592243129584e-05,
+      "loss": 3.7696,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3386206295619064e-05,
+      "loss": 3.7653,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337783672691227e-05,
+      "loss": 3.7741,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336945077940175e-05,
+      "loss": 3.7726,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336106483189123e-05,
+      "loss": 3.7696,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335267888438071e-05,
+      "loss": 3.7615,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334430931567393e-05,
+      "loss": 3.7496,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333592336816341e-05,
+      "loss": 3.777,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332753742065289e-05,
+      "loss": 3.7691,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331915147314237e-05,
+      "loss": 3.7622,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331078190443558e-05,
+      "loss": 3.7613,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330239595692506e-05,
+      "loss": 3.766,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329401000941454e-05,
+      "loss": 3.7507,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328562406190402e-05,
+      "loss": 3.7671,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3277254493197226e-05,
+      "loss": 3.7661,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268868545686706e-05,
+      "loss": 3.7675,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3260482598176186e-05,
+      "loss": 3.7617,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3252096650665666e-05,
+      "loss": 3.7645,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324372708195888e-05,
+      "loss": 3.7564,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323534113444836e-05,
+      "loss": 3.7681,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322695518693784e-05,
+      "loss": 3.7518,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321856923942732e-05,
+      "loss": 3.757,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321019967072053e-05,
+      "loss": 3.7723,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320181372321001e-05,
+      "loss": 3.7597,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319342777569949e-05,
+      "loss": 3.7532,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318504182818897e-05,
+      "loss": 3.7552,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317665588067845e-05,
+      "loss": 3.7541,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316828631197166e-05,
+      "loss": 3.7665,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315990036446114e-05,
+      "loss": 3.7698,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315151441695062e-05,
+      "loss": 3.7638,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31431284694401e-05,
+      "loss": 3.7739,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134758900733316e-05,
+      "loss": 3.7708,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3126372953222796e-05,
+      "loss": 3.7663,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117987005712276e-05,
+      "loss": 3.7609,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109601058201756e-05,
+      "loss": 3.7629,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101231489494965e-05,
+      "loss": 3.7585,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092845541984445e-05,
+      "loss": 3.7526,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084459594473925e-05,
+      "loss": 3.7644,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3076073646963405e-05,
+      "loss": 3.7566,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067704078256614e-05,
+      "loss": 3.769,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3059318130746094e-05,
+      "loss": 3.7702,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050932183235574e-05,
+      "loss": 3.743,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042546235725054e-05,
+      "loss": 3.7624,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303417666701827e-05,
+      "loss": 3.7567,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302579071950775e-05,
+      "loss": 3.7628,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301740477199723e-05,
+      "loss": 3.7567,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300901882448671e-05,
+      "loss": 3.7479,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300064925577992e-05,
+      "loss": 3.7607,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29922633082694e-05,
+      "loss": 3.7525,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298387736075888e-05,
+      "loss": 3.7568,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297550779205209e-05,
+      "loss": 3.7512,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296712184454157e-05,
+      "loss": 3.7618,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295873589703105e-05,
+      "loss": 3.7615,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295034994952053e-05,
+      "loss": 3.7609,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294198038081374e-05,
+      "loss": 3.7448,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293359443330322e-05,
+      "loss": 3.758,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29252084857927e-05,
+      "loss": 3.7569,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291682253828218e-05,
+      "loss": 3.7661,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290843659077166e-05,
+      "loss": 3.7503,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290006702206487e-05,
+      "loss": 3.7534,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289168107455435e-05,
+      "loss": 3.7505,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288329512704383e-05,
+      "loss": 3.7633,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287490917953331e-05,
+      "loss": 3.7444,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286653961082652e-05,
+      "loss": 3.749,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858153663316e-05,
+      "loss": 3.7517,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284976771580548e-05,
+      "loss": 3.7516,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28413981470987e-05,
+      "loss": 3.7422,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283301219958818e-05,
+      "loss": 3.7508,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282462625207766e-05,
+      "loss": 3.7472,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281624030456714e-05,
+      "loss": 3.7542,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807854357056617e-05,
+      "loss": 3.7537,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799468409546097e-05,
+      "loss": 3.7432,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2791082462035576e-05,
+      "loss": 3.7564,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278269651452505e-05,
+      "loss": 3.7669,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2774326945818266e-05,
+      "loss": 3.7606,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2765940998307746e-05,
+      "loss": 3.7406,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2757555050797225e-05,
+      "loss": 3.7507,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27491691032867e-05,
+      "loss": 3.7489,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2740799534579915e-05,
+      "loss": 3.7451,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732413587069394e-05,
+      "loss": 3.7492,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2724027639558874e-05,
+      "loss": 3.7398,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715641692048354e-05,
+      "loss": 3.7518,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270727212334157e-05,
+      "loss": 3.7465,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269888617583105e-05,
+      "loss": 3.7575,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690500228320523e-05,
+      "loss": 3.745,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268211428081e-05,
+      "loss": 3.7424,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267374471210322e-05,
+      "loss": 3.753,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26653587645927e-05,
+      "loss": 3.7578,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265697281708217e-05,
+      "loss": 3.7508,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264858686957165e-05,
+      "loss": 3.7518,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264021730086487e-05,
+      "loss": 3.7407,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263183135335435e-05,
+      "loss": 3.7562,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262344540584383e-05,
+      "loss": 3.7489,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261505945833331e-05,
+      "loss": 3.7578,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606689889626524e-05,
+      "loss": 3.7419,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2598303942116e-05,
+      "loss": 3.756,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258991799460548e-05,
+      "loss": 3.7452,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258153204709496e-05,
+      "loss": 3.7676,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257316247838817e-05,
+      "loss": 3.7402,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564776530877646e-05,
+      "loss": 3.7566,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2556390583367126e-05,
+      "loss": 3.7458,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2548004635856606e-05,
+      "loss": 3.742,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253963506714982e-05,
+      "loss": 3.7534,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25312491196393e-05,
+      "loss": 3.7495,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252286317212878e-05,
+      "loss": 3.7519,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251447722461826e-05,
+      "loss": 3.7409,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250610765591147e-05,
+      "loss": 3.7493,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.870181083679199,
+      "eval_runtime": 305.8525,
+      "eval_samples_per_second": 1247.631,
+      "eval_steps_per_second": 38.989,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249772170840095e-05,
+      "loss": 3.7411,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248933576089043e-05,
+      "loss": 3.7343,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248094981337991e-05,
+      "loss": 3.7432,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247258024467312e-05,
+      "loss": 3.7466,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24641942971626e-05,
+      "loss": 3.7533,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245580834965208e-05,
+      "loss": 3.7429,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244742240214156e-05,
+      "loss": 3.7327,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2439052833434776e-05,
+      "loss": 3.7406,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2430666885924255e-05,
+      "loss": 3.7509,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2422280938413735e-05,
+      "loss": 3.7409,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2413894990903215e-05,
+      "loss": 3.7365,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2405525422196424e-05,
+      "loss": 3.7357,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239715585348964e-05,
+      "loss": 3.7414,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2388769905979114e-05,
+      "loss": 3.7354,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2380383958468593e-05,
+      "loss": 3.7419,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2371998010958073e-05,
+      "loss": 3.7276,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2363612063447553e-05,
+      "loss": 3.7273,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235522611593703e-05,
+      "loss": 3.7307,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234684016842651e-05,
+      "loss": 3.7264,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233845422091599e-05,
+      "loss": 3.7482,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233008465220921e-05,
+      "loss": 3.7428,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232169870469869e-05,
+      "loss": 3.7359,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231331275718817e-05,
+      "loss": 3.7323,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230492680967765e-05,
+      "loss": 3.7412,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229655724097086e-05,
+      "loss": 3.73,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228817129346034e-05,
+      "loss": 3.7401,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227978534594982e-05,
+      "loss": 3.7286,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.22713993984393e-05,
+      "loss": 3.7336,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226302982973251e-05,
+      "loss": 3.7323,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225464388222199e-05,
+      "loss": 3.7294,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224625793471147e-05,
+      "loss": 3.7344,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223787198720095e-05,
+      "loss": 3.7395,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222950241849416e-05,
+      "loss": 3.738,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222111647098364e-05,
+      "loss": 3.7423,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221273052347312e-05,
+      "loss": 3.7262,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.22043445759626e-05,
+      "loss": 3.7419,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219597500725581e-05,
+      "loss": 3.7316,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218758905974529e-05,
+      "loss": 3.7235,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217920311223477e-05,
+      "loss": 3.726,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217081716472425e-05,
+      "loss": 3.7262,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216244759601746e-05,
+      "loss": 3.7294,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215406164850694e-05,
+      "loss": 3.7379,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214567570099642e-05,
+      "loss": 3.7288,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21372897534859e-05,
+      "loss": 3.7298,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2128920184779116e-05,
+      "loss": 3.7352,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2120534237268596e-05,
+      "loss": 3.7316,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2112148289758076e-05,
+      "loss": 3.7318,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2103762342247556e-05,
+      "loss": 3.7194,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2095392773540765e-05,
+      "loss": 3.7102,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2087006826030245e-05,
+      "loss": 3.7354,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2078620878519725e-05,
+      "loss": 3.7345,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2070234931009205e-05,
+      "loss": 3.7203,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061865362302414e-05,
+      "loss": 3.7286,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2053479414791894e-05,
+      "loss": 3.7255,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2045093467281374e-05,
+      "loss": 3.7125,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036707519770854e-05,
+      "loss": 3.7244,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.202833795106407e-05,
+      "loss": 3.7294,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201995200355355e-05,
+      "loss": 3.7289,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201156605604303e-05,
+      "loss": 3.7212,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200318010853251e-05,
+      "loss": 3.7253,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199481053982572e-05,
+      "loss": 3.7184,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19864245923152e-05,
+      "loss": 3.7298,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197803864480468e-05,
+      "loss": 3.7138,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196965269729416e-05,
+      "loss": 3.7169,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196128312858737e-05,
+      "loss": 3.732,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195289718107685e-05,
+      "loss": 3.7209,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194451123356633e-05,
+      "loss": 3.7179,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193612528605581e-05,
+      "loss": 3.7157,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1927755717349024e-05,
+      "loss": 3.7137,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1919369769838504e-05,
+      "loss": 3.7219,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1910983822327984e-05,
+      "loss": 3.7347,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1902597874817464e-05,
+      "loss": 3.7233,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.189422830611067e-05,
+      "loss": 3.7363,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188584235860015e-05,
+      "loss": 3.7301,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187745641108963e-05,
+      "loss": 3.7314,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186907046357911e-05,
+      "loss": 3.723,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186070089487232e-05,
+      "loss": 3.7226,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18523149473618e-05,
+      "loss": 3.7209,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184392899985128e-05,
+      "loss": 3.7112,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.183554305234076e-05,
+      "loss": 3.7255,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182717348363398e-05,
+      "loss": 3.7203,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181878753612346e-05,
+      "loss": 3.7313,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181040158861294e-05,
+      "loss": 3.7281,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1802032019906146e-05,
+      "loss": 3.7039,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1793646072395626e-05,
+      "loss": 3.7283,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1785260124885106e-05,
+      "loss": 3.7189,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776874177374586e-05,
+      "loss": 3.7224,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1768504608667795e-05,
+      "loss": 3.7176,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1760118661157275e-05,
+      "loss": 3.7097,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1751732713646755e-05,
+      "loss": 3.7226,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1743346766136235e-05,
+      "loss": 3.7128,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173497719742945e-05,
+      "loss": 3.7183,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172659124991893e-05,
+      "loss": 3.7154,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171820530240841e-05,
+      "loss": 3.7232,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170981935489789e-05,
+      "loss": 3.7288,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17014497861911e-05,
+      "loss": 3.7152,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169306383868058e-05,
+      "loss": 3.7126,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168467789117006e-05,
+      "loss": 3.7198,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167629194365954e-05,
+      "loss": 3.7174,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166792237495275e-05,
+      "loss": 3.7291,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165953642744223e-05,
+      "loss": 3.7128,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165115047993171e-05,
+      "loss": 3.7173,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164276453242119e-05,
+      "loss": 3.7131,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1634394963714405e-05,
+      "loss": 3.7252,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1626009016203885e-05,
+      "loss": 3.7045,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1617623068693365e-05,
+      "loss": 3.7129,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1609237121182845e-05,
+      "loss": 3.7144,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1600867552476054e-05,
+      "loss": 3.7145,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159249798376926e-05,
+      "loss": 3.7043,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158411203625874e-05,
+      "loss": 3.7149,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157572608874822e-05,
+      "loss": 3.7076,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15673401412377e-05,
+      "loss": 3.7196,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155895419372718e-05,
+      "loss": 3.7151,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155056824621666e-05,
+      "loss": 3.7098,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154218229870614e-05,
+      "loss": 3.7166,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153379635119562e-05,
+      "loss": 3.729,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152542678248884e-05,
+      "loss": 3.7273,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151704083497832e-05,
+      "loss": 3.7074,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150865488746779e-05,
+      "loss": 3.7137,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150026893995727e-05,
+      "loss": 3.7149,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149189937125049e-05,
+      "loss": 3.7056,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.148351342373997e-05,
+      "loss": 3.7153,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147512747622944e-05,
+      "loss": 3.699,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146674152871892e-05,
+      "loss": 3.7196,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1458371960012136e-05,
+      "loss": 3.7061,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144998601250161e-05,
+      "loss": 3.7231,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1441600064991096e-05,
+      "loss": 3.7093,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1433214117480576e-05,
+      "loss": 3.7084,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142484454877379e-05,
+      "loss": 3.7128,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1416458601263265e-05,
+      "loss": 3.7226,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1408072653752745e-05,
+      "loss": 3.7109,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1399686706242225e-05,
+      "loss": 3.7187,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139131713753544e-05,
+      "loss": 3.703,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382931190024914e-05,
+      "loss": 3.7209,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1374545242514394e-05,
+      "loss": 3.7116,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1366159295003874e-05,
+      "loss": 3.7169,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.135778972629708e-05,
+      "loss": 3.7101,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134940377878656e-05,
+      "loss": 3.7199,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134101783127604e-05,
+      "loss": 3.7114,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133263188376553e-05,
+      "loss": 3.7298,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132426231505874e-05,
+      "loss": 3.7038,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131587636754822e-05,
+      "loss": 3.7181,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13074904200377e-05,
+      "loss": 3.7104,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129910447252718e-05,
+      "loss": 3.7106,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129073490382039e-05,
+      "loss": 3.7087,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128234895630987e-05,
+      "loss": 3.7154,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127396300879935e-05,
+      "loss": 3.7147,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126557706128883e-05,
+      "loss": 3.7064,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125720749258204e-05,
+      "loss": 3.7175,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.860360622406006,
+      "eval_runtime": 304.7877,
+      "eval_samples_per_second": 1251.989,
+      "eval_steps_per_second": 39.126,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124882154507152e-05,
+      "loss": 3.7075,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1240435597561e-05,
+      "loss": 3.6998,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1232049650050484e-05,
+      "loss": 3.7073,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1223663702539964e-05,
+      "loss": 3.7095,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1215277755029444e-05,
+      "loss": 3.716,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1206891807518923e-05,
+      "loss": 3.7095,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1198505860008403e-05,
+      "loss": 3.7027,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119011991249788e-05,
+      "loss": 3.7032,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118175034379109e-05,
+      "loss": 3.717,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117336439628057e-05,
+      "loss": 3.7065,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116497844877005e-05,
+      "loss": 3.7017,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115659250125953e-05,
+      "loss": 3.6998,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114823931135647e-05,
+      "loss": 3.7076,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113985336384595e-05,
+      "loss": 3.6999,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113146741633544e-05,
+      "loss": 3.7054,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112308146882492e-05,
+      "loss": 3.6909,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11146955213144e-05,
+      "loss": 3.6952,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110630957380388e-05,
+      "loss": 3.6981,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109792362629336e-05,
+      "loss": 3.6921,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108953767878284e-05,
+      "loss": 3.7117,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1081168110076046e-05,
+      "loss": 3.7096,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072782162565526e-05,
+      "loss": 3.7012,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1064396215055006e-05,
+      "loss": 3.6983,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1056010267544486e-05,
+      "loss": 3.708,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1047640698837695e-05,
+      "loss": 3.6951,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1039254751327175e-05,
+      "loss": 3.7037,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030868803816655e-05,
+      "loss": 3.692,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1022482856306135e-05,
+      "loss": 3.7029,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101411328759935e-05,
+      "loss": 3.7021,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100572734008883e-05,
+      "loss": 3.6929,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099734139257831e-05,
+      "loss": 3.7002,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098895544506779e-05,
+      "loss": 3.7059,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0980585876361e-05,
+      "loss": 3.7011,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097219992885048e-05,
+      "loss": 3.7121,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096381398133996e-05,
+      "loss": 3.6926,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095544441263317e-05,
+      "loss": 3.7022,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094705846512265e-05,
+      "loss": 3.6996,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093867251761213e-05,
+      "loss": 3.6909,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093028657010161e-05,
+      "loss": 3.6915,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921917001394825e-05,
+      "loss": 3.692,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913531053884305e-05,
+      "loss": 3.6927,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0905145106373784e-05,
+      "loss": 3.7027,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896759158863264e-05,
+      "loss": 3.6967,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0888389590156474e-05,
+      "loss": 3.6963,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0880003642645953e-05,
+      "loss": 3.7009,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0871617695135433e-05,
+      "loss": 3.6959,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0863231747624913e-05,
+      "loss": 3.6953,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085486217891812e-05,
+      "loss": 3.689,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08464762314076e-05,
+      "loss": 3.6701,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083809028389708e-05,
+      "loss": 3.7091,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082970433638656e-05,
+      "loss": 3.6946,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082133476767978e-05,
+      "loss": 3.6922,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081294882016926e-05,
+      "loss": 3.6934,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080456287265874e-05,
+      "loss": 3.6935,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079617692514822e-05,
+      "loss": 3.6799,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078780735644143e-05,
+      "loss": 3.6882,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077942140893091e-05,
+      "loss": 3.6967,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077103546142039e-05,
+      "loss": 3.6969,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076264951390987e-05,
+      "loss": 3.694,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0754279945203076e-05,
+      "loss": 3.6863,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745893997692556e-05,
+      "loss": 3.6862,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0737508050182036e-05,
+      "loss": 3.7006,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0729122102671516e-05,
+      "loss": 3.6754,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072075253396473e-05,
+      "loss": 3.684,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071236658645421e-05,
+      "loss": 3.6967,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070398063894369e-05,
+      "loss": 3.6913,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0695594691433165e-05,
+      "loss": 3.6823,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068722512272638e-05,
+      "loss": 3.6874,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067883917521586e-05,
+      "loss": 3.6797,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067045322770534e-05,
+      "loss": 3.6888,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0662067280194814e-05,
+      "loss": 3.7012,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065369771148803e-05,
+      "loss": 3.69,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064531176397751e-05,
+      "loss": 3.7034,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063692581646699e-05,
+      "loss": 3.6976,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062853986895647e-05,
+      "loss": 3.702,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0620170300249686e-05,
+      "loss": 3.6885,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0611784352739166e-05,
+      "loss": 3.6938,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060339840522864e-05,
+      "loss": 3.6873,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059501245771812e-05,
+      "loss": 3.6755,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0586642889011335e-05,
+      "loss": 3.6932,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0578256941500814e-05,
+      "loss": 3.688,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056987099399029e-05,
+      "loss": 3.6973,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056148504647977e-05,
+      "loss": 3.6987,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0553115477772983e-05,
+      "loss": 3.6761,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054472953026246e-05,
+      "loss": 3.6908,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053634358275194e-05,
+      "loss": 3.6859,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052795763524142e-05,
+      "loss": 3.6931,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051958806653464e-05,
+      "loss": 3.6845,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051120211902411e-05,
+      "loss": 3.6813,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050281617151359e-05,
+      "loss": 3.6866,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049443022400307e-05,
+      "loss": 3.682,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048606065529629e-05,
+      "loss": 3.6867,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047767470778576e-05,
+      "loss": 3.6842,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046928876027524e-05,
+      "loss": 3.6953,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046090281276472e-05,
+      "loss": 3.6929,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045253324405793e-05,
+      "loss": 3.6832,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044414729654741e-05,
+      "loss": 3.6767,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043576134903689e-05,
+      "loss": 3.6874,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042737540152638e-05,
+      "loss": 3.6913,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0419005832819586e-05,
+      "loss": 3.6935,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0410619885309066e-05,
+      "loss": 3.6854,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0402233937798546e-05,
+      "loss": 3.6846,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039386436909176e-05,
+      "loss": 3.6801,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0385478421581235e-05,
+      "loss": 3.697,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0377092474070715e-05,
+      "loss": 3.6727,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0368706526560195e-05,
+      "loss": 3.6834,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0360336957853404e-05,
+      "loss": 3.6771,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0351951010342884e-05,
+      "loss": 3.6856,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03435814416361e-05,
+      "loss": 3.6731,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033519549412558e-05,
+      "loss": 3.6881,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032680954661506e-05,
+      "loss": 3.6738,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031842359910454e-05,
+      "loss": 3.689,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031003765159402e-05,
+      "loss": 3.6842,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03016517040835e-05,
+      "loss": 3.6766,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029326575657298e-05,
+      "loss": 3.6857,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028487980906246e-05,
+      "loss": 3.7,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027651024035567e-05,
+      "loss": 3.6871,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026812429284515e-05,
+      "loss": 3.6857,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025973834533463e-05,
+      "loss": 3.6751,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025135239782411e-05,
+      "loss": 3.6832,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024298282911732e-05,
+      "loss": 3.6822,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02345968816068e-05,
+      "loss": 3.6791,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0226210934096284e-05,
+      "loss": 3.6679,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217824986585764e-05,
+      "loss": 3.6868,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0209455417878973e-05,
+      "loss": 3.6779,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020106947036845e-05,
+      "loss": 3.6891,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019268352285793e-05,
+      "loss": 3.6785,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018429757534741e-05,
+      "loss": 3.6791,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017592800664062e-05,
+      "loss": 3.6796,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01675420591301e-05,
+      "loss": 3.6926,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015915611161958e-05,
+      "loss": 3.6803,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015077016410906e-05,
+      "loss": 3.6872,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014240059540227e-05,
+      "loss": 3.6685,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013401464789175e-05,
+      "loss": 3.6925,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012562870038124e-05,
+      "loss": 3.6809,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011724275287072e-05,
+      "loss": 3.6876,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010887318416393e-05,
+      "loss": 3.681,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010048723665341e-05,
+      "loss": 3.6871,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009210128914289e-05,
+      "loss": 3.6788,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008371534163237e-05,
+      "loss": 3.6983,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0075345772925576e-05,
+      "loss": 3.6746,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0066959825415056e-05,
+      "loss": 3.6907,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0058573877904536e-05,
+      "loss": 3.6833,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0050187930394016e-05,
+      "loss": 3.6779,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0041818361687225e-05,
+      "loss": 3.6762,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0033432414176705e-05,
+      "loss": 3.6864,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002504646666619e-05,
+      "loss": 3.6846,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00166768979594e-05,
+      "loss": 3.6759,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000829095044888e-05,
+      "loss": 3.6858,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8530972003936768,
+      "eval_runtime": 304.6684,
+      "eval_samples_per_second": 1252.48,
+      "eval_steps_per_second": 39.141,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999990500293836e-05,
+      "loss": 3.669,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999151905542784e-05,
+      "loss": 3.6714,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998314948672105e-05,
+      "loss": 3.6755,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997476353921053e-05,
+      "loss": 3.6807,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996637759170001e-05,
+      "loss": 3.6878,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995799164418949e-05,
+      "loss": 3.6748,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99496220754827e-05,
+      "loss": 3.6776,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994123612797218e-05,
+      "loss": 3.6723,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.993285018046166e-05,
+      "loss": 3.6844,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9924464232951145e-05,
+      "loss": 3.682,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9916094664244354e-05,
+      "loss": 3.6698,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9907708716733834e-05,
+      "loss": 3.668,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9899339148027044e-05,
+      "loss": 3.6802,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9890953200516523e-05,
+      "loss": 3.6685,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9882567253006003e-05,
+      "loss": 3.6723,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987418130549548e-05,
+      "loss": 3.6657,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986579535798496e-05,
+      "loss": 3.6625,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.985740941047444e-05,
+      "loss": 3.6712,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984902346296392e-05,
+      "loss": 3.6614,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98406375154534e-05,
+      "loss": 3.6806,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983226794674661e-05,
+      "loss": 3.68,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98238819992361e-05,
+      "loss": 3.6755,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981549605172558e-05,
+      "loss": 3.6676,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.980711010421506e-05,
+      "loss": 3.6716,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979874053550827e-05,
+      "loss": 3.6708,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979035458799775e-05,
+      "loss": 3.6746,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.978196864048723e-05,
+      "loss": 3.6666,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.977358269297671e-05,
+      "loss": 3.6732,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976521312426992e-05,
+      "loss": 3.6678,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.97568271767594e-05,
+      "loss": 3.6651,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974844122924888e-05,
+      "loss": 3.667,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974005528173836e-05,
+      "loss": 3.6794,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9731685713031566e-05,
+      "loss": 3.6761,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9723299765521046e-05,
+      "loss": 3.6803,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.971491381801053e-05,
+      "loss": 3.6633,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970652787050001e-05,
+      "loss": 3.6681,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969815830179322e-05,
+      "loss": 3.6763,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96897723542827e-05,
+      "loss": 3.6624,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968138640677218e-05,
+      "loss": 3.6605,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967300045926166e-05,
+      "loss": 3.6619,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966463089055487e-05,
+      "loss": 3.6652,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965624494304435e-05,
+      "loss": 3.6672,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964785899553383e-05,
+      "loss": 3.6751,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963948942682704e-05,
+      "loss": 3.6694,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963110347931652e-05,
+      "loss": 3.6651,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9622717531806e-05,
+      "loss": 3.6714,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9614331584295486e-05,
+      "loss": 3.6637,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9605962015588695e-05,
+      "loss": 3.664,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9597576068078175e-05,
+      "loss": 3.6395,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9589190120567655e-05,
+      "loss": 3.6816,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9580804173057135e-05,
+      "loss": 3.6641,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957241822554661e-05,
+      "loss": 3.6719,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956403227803609e-05,
+      "loss": 3.6624,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9555662709329304e-05,
+      "loss": 3.6631,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9547276761818784e-05,
+      "loss": 3.65,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953889081430826e-05,
+      "loss": 3.6581,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953050486679774e-05,
+      "loss": 3.6702,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9522118919287224e-05,
+      "loss": 3.6639,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9513732971776704e-05,
+      "loss": 3.671,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9505347024266184e-05,
+      "loss": 3.6609,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949697745555939e-05,
+      "loss": 3.6536,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948859150804887e-05,
+      "loss": 3.6733,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948020556053835e-05,
+      "loss": 3.6483,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947181961302783e-05,
+      "loss": 3.6518,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946343366551731e-05,
+      "loss": 3.6682,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945504771800679e-05,
+      "loss": 3.6675,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944666177049627e-05,
+      "loss": 3.6507,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943827582298575e-05,
+      "loss": 3.6654,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942990625427896e-05,
+      "loss": 3.6459,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942152030676844e-05,
+      "loss": 3.6647,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941313435925792e-05,
+      "loss": 3.669,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940474841174741e-05,
+      "loss": 3.6663,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939637884304062e-05,
+      "loss": 3.6699,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.93879928955301e-05,
+      "loss": 3.6706,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937960694801958e-05,
+      "loss": 3.6779,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937122100050906e-05,
+      "loss": 3.6581,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9362851431802267e-05,
+      "loss": 3.6658,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9354465484291747e-05,
+      "loss": 3.6619,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9346079536781227e-05,
+      "loss": 3.6438,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9337693589270706e-05,
+      "loss": 3.6662,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9329324020563916e-05,
+      "loss": 3.6637,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9320938073053396e-05,
+      "loss": 3.6659,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9312552125542875e-05,
+      "loss": 3.6754,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930416617803236e-05,
+      "loss": 3.6504,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929579660932557e-05,
+      "loss": 3.6562,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928741066181505e-05,
+      "loss": 3.659,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927902471430453e-05,
+      "loss": 3.6643,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927065514559774e-05,
+      "loss": 3.6526,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926226919808722e-05,
+      "loss": 3.6564,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.92538832505767e-05,
+      "loss": 3.6593,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924549730306618e-05,
+      "loss": 3.6565,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923712773435939e-05,
+      "loss": 3.6551,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922874178684887e-05,
+      "loss": 3.6609,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922035583933835e-05,
+      "loss": 3.6676,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921196989182783e-05,
+      "loss": 3.6598,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9203600323121045e-05,
+      "loss": 3.6656,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9195214375610525e-05,
+      "loss": 3.6474,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9186828428100005e-05,
+      "loss": 3.6575,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9178442480589485e-05,
+      "loss": 3.6614,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9170072911882694e-05,
+      "loss": 3.6666,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9161686964372174e-05,
+      "loss": 3.6584,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9153301016861654e-05,
+      "loss": 3.6534,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9144915069351134e-05,
+      "loss": 3.6587,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913654550064434e-05,
+      "loss": 3.6615,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912815955313382e-05,
+      "loss": 3.6488,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91197736056233e-05,
+      "loss": 3.6566,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911138765811278e-05,
+      "loss": 3.6436,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9103018089406e-05,
+      "loss": 3.6655,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909464852069921e-05,
+      "loss": 3.65,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908626257318869e-05,
+      "loss": 3.6566,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.907787662567817e-05,
+      "loss": 3.6469,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906949067816765e-05,
+      "loss": 3.6592,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906110473065713e-05,
+      "loss": 3.6599,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905271878314661e-05,
+      "loss": 3.6507,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904433283563609e-05,
+      "loss": 3.6565,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903594688812556e-05,
+      "loss": 3.6687,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9027577319418777e-05,
+      "loss": 3.6643,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9019191371908257e-05,
+      "loss": 3.6535,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9010805424397736e-05,
+      "loss": 3.65,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9002419476887216e-05,
+      "loss": 3.6538,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.899404990818043e-05,
+      "loss": 3.6588,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898566396066991e-05,
+      "loss": 3.6477,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.897727801315939e-05,
+      "loss": 3.6447,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8968892065648865e-05,
+      "loss": 3.66,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.896052249694208e-05,
+      "loss": 3.6539,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895213654943156e-05,
+      "loss": 3.6613,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8943750601921034e-05,
+      "loss": 3.6523,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8935364654410514e-05,
+      "loss": 3.6536,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.892699508570373e-05,
+      "loss": 3.6501,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891860913819321e-05,
+      "loss": 3.6654,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891022319068269e-05,
+      "loss": 3.6569,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890183724317217e-05,
+      "loss": 3.6572,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8893467674465386e-05,
+      "loss": 3.6445,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8885081726954866e-05,
+      "loss": 3.6667,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887669577944434e-05,
+      "loss": 3.6549,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886830983193382e-05,
+      "loss": 3.6623,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8859940263227035e-05,
+      "loss": 3.6547,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885155431571651e-05,
+      "loss": 3.6579,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884316836820599e-05,
+      "loss": 3.6543,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883478242069547e-05,
+      "loss": 3.6728,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8826412851988684e-05,
+      "loss": 3.6481,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8818026904478164e-05,
+      "loss": 3.6635,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8809640956967644e-05,
+      "loss": 3.6567,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8801255009457124e-05,
+      "loss": 3.6489,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879288544075034e-05,
+      "loss": 3.653,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878449949323981e-05,
+      "loss": 3.6592,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877611354572929e-05,
+      "loss": 3.6572,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876774397702251e-05,
+      "loss": 3.6545,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.875935802951198e-05,
+      "loss": 3.6577,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8489201068878174,
+      "eval_runtime": 306.5637,
+      "eval_samples_per_second": 1244.736,
+      "eval_steps_per_second": 38.899,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.875097208200146e-05,
+      "loss": 3.6505,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874258613449094e-05,
+      "loss": 3.6436,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873420018698042e-05,
+      "loss": 3.6501,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87258142394699e-05,
+      "loss": 3.6554,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871742829195938e-05,
+      "loss": 3.6616,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870904234444886e-05,
+      "loss": 3.6479,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870065639693835e-05,
+      "loss": 3.6519,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869227044942783e-05,
+      "loss": 3.6482,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868390088072104e-05,
+      "loss": 3.6513,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867551493321052e-05,
+      "loss": 3.6611,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86671289857e-05,
+      "loss": 3.6421,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865874303818948e-05,
+      "loss": 3.6411,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8650389848286415e-05,
+      "loss": 3.6595,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8642003900775895e-05,
+      "loss": 3.6401,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8633617953265375e-05,
+      "loss": 3.6453,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8625232005754855e-05,
+      "loss": 3.6373,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861686243704807e-05,
+      "loss": 3.6382,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.860847648953755e-05,
+      "loss": 3.6418,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.860009054202703e-05,
+      "loss": 3.6362,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859170459451651e-05,
+      "loss": 3.6512,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858331864700599e-05,
+      "loss": 3.6523,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857493269949547e-05,
+      "loss": 3.6527,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856654675198495e-05,
+      "loss": 3.6426,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855816080447443e-05,
+      "loss": 3.6449,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854979123576764e-05,
+      "loss": 3.6482,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854140528825712e-05,
+      "loss": 3.6444,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85330193407466e-05,
+      "loss": 3.641,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852464977203981e-05,
+      "loss": 3.6551,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851626382452929e-05,
+      "loss": 3.6376,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850787787701877e-05,
+      "loss": 3.6396,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8499491929508256e-05,
+      "loss": 3.64,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8491105981997736e-05,
+      "loss": 3.652,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8482720034487216e-05,
+      "loss": 3.6503,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8474334086976695e-05,
+      "loss": 3.6608,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8465964518269905e-05,
+      "loss": 3.6337,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8457578570759385e-05,
+      "loss": 3.644,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8449192623248865e-05,
+      "loss": 3.653,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8440806675738344e-05,
+      "loss": 3.6313,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843245348583528e-05,
+      "loss": 3.6379,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842406753832476e-05,
+      "loss": 3.6375,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841568159081424e-05,
+      "loss": 3.6403,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840729564330372e-05,
+      "loss": 3.6422,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839890969579321e-05,
+      "loss": 3.6455,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839052374828269e-05,
+      "loss": 3.647,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838213780077217e-05,
+      "loss": 3.6396,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837375185326164e-05,
+      "loss": 3.6458,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836538228455486e-05,
+      "loss": 3.6397,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835701271584807e-05,
+      "loss": 3.6394,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834862676833755e-05,
+      "loss": 3.6116,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834024082082703e-05,
+      "loss": 3.6554,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833185487331651e-05,
+      "loss": 3.6375,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832346892580599e-05,
+      "loss": 3.6455,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831508297829547e-05,
+      "loss": 3.6381,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8306713409588676e-05,
+      "loss": 3.6417,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829832746207816e-05,
+      "loss": 3.6256,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828994151456764e-05,
+      "loss": 3.6335,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8281555567057116e-05,
+      "loss": 3.6499,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8273169619546596e-05,
+      "loss": 3.6345,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8264783672036076e-05,
+      "loss": 3.6448,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825641410332929e-05,
+      "loss": 3.6395,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8248028155818765e-05,
+      "loss": 3.6262,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8239642208308245e-05,
+      "loss": 3.6512,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8231256260797725e-05,
+      "loss": 3.624,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8222870313287205e-05,
+      "loss": 3.6301,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8214484365776685e-05,
+      "loss": 3.6447,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8206098418266165e-05,
+      "loss": 3.6391,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8197712470755645e-05,
+      "loss": 3.6281,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818934290204886e-05,
+      "loss": 3.6424,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818095695453834e-05,
+      "loss": 3.6217,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817257100702782e-05,
+      "loss": 3.6415,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81641850595173e-05,
+      "loss": 3.6452,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.815581549081051e-05,
+      "loss": 3.6376,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814744592210372e-05,
+      "loss": 3.6452,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81390599745932e-05,
+      "loss": 3.6464,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813067402708268e-05,
+      "loss": 3.6524,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812228807957216e-05,
+      "loss": 3.633,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811391851086537e-05,
+      "loss": 3.6384,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810553256335485e-05,
+      "loss": 3.6424,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8097146615844334e-05,
+      "loss": 3.6162,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8088760668333814e-05,
+      "loss": 3.6409,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8080374720823294e-05,
+      "loss": 3.6421,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8071988773312774e-05,
+      "loss": 3.639,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8063602825802254e-05,
+      "loss": 3.6498,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805523325709546e-05,
+      "loss": 3.6267,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804684730958494e-05,
+      "loss": 3.6303,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803846136207442e-05,
+      "loss": 3.6334,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80300754145639e-05,
+      "loss": 3.6386,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802168946705338e-05,
+      "loss": 3.634,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801330351954286e-05,
+      "loss": 3.6312,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800493395083607e-05,
+      "loss": 3.6293,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799654800332555e-05,
+      "loss": 3.635,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798816205581503e-05,
+      "loss": 3.6313,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797977610830452e-05,
+      "loss": 3.6355,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7971390160794e-05,
+      "loss": 3.6449,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796300421328348e-05,
+      "loss": 3.6354,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795461826577295e-05,
+      "loss": 3.6409,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794623231826243e-05,
+      "loss": 3.6248,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793786274955565e-05,
+      "loss": 3.6337,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792947680204513e-05,
+      "loss": 3.6352,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79210908545346e-05,
+      "loss": 3.6437,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.791270490702408e-05,
+      "loss": 3.6334,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79043353383173e-05,
+      "loss": 3.6321,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7895965769610506e-05,
+      "loss": 3.6352,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7887579822099986e-05,
+      "loss": 3.6371,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787919387458947e-05,
+      "loss": 3.6234,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787080792707895e-05,
+      "loss": 3.6345,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7862421979568426e-05,
+      "loss": 3.6211,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7854036032057906e-05,
+      "loss": 3.6396,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784566646335112e-05,
+      "loss": 3.6278,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78372805158406e-05,
+      "loss": 3.6298,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7828894568330075e-05,
+      "loss": 3.6216,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7820508620819555e-05,
+      "loss": 3.6409,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7812122673309034e-05,
+      "loss": 3.6325,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7803736725798514e-05,
+      "loss": 3.6258,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7795350778287994e-05,
+      "loss": 3.6356,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7786964830777474e-05,
+      "loss": 3.6437,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777859526207069e-05,
+      "loss": 3.6409,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777020931456017e-05,
+      "loss": 3.6295,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776183974585338e-05,
+      "loss": 3.6246,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775345379834286e-05,
+      "loss": 3.6327,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774506785083234e-05,
+      "loss": 3.6356,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773668190332182e-05,
+      "loss": 3.6213,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.77282959558113e-05,
+      "loss": 3.6215,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771991000830078e-05,
+      "loss": 3.6368,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771152406079026e-05,
+      "loss": 3.6331,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770315449208347e-05,
+      "loss": 3.6326,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769476854457295e-05,
+      "loss": 3.6292,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768638259706243e-05,
+      "loss": 3.6297,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767799664955191e-05,
+      "loss": 3.6279,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766961070204139e-05,
+      "loss": 3.6407,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7661224754530875e-05,
+      "loss": 3.6295,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7652838807020355e-05,
+      "loss": 3.6379,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7644469238313564e-05,
+      "loss": 3.619,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7636083290803044e-05,
+      "loss": 3.6443,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7627697343292524e-05,
+      "loss": 3.6251,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7619311395782004e-05,
+      "loss": 3.6414,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7610925448271484e-05,
+      "loss": 3.6301,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7602539500760963e-05,
+      "loss": 3.6357,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759416993205417e-05,
+      "loss": 3.6267,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758578398454365e-05,
+      "loss": 3.6516,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757739803703313e-05,
+      "loss": 3.6255,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756901208952261e-05,
+      "loss": 3.6391,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756062614201209e-05,
+      "loss": 3.6352,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755224019450157e-05,
+      "loss": 3.6236,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754385424699105e-05,
+      "loss": 3.6319,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7535501057088e-05,
+      "loss": 3.6326,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752711510957748e-05,
+      "loss": 3.6338,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751872916206696e-05,
+      "loss": 3.6308,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751034321455644e-05,
+      "loss": 3.6345,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.846997022628784,
+      "eval_runtime": 308.8167,
+      "eval_samples_per_second": 1235.655,
+      "eval_steps_per_second": 38.615,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750195726704591e-05,
+      "loss": 3.6201,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749357131953539e-05,
+      "loss": 3.6206,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748518537202487e-05,
+      "loss": 3.6236,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.747679942451435e-05,
+      "loss": 3.6321,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7468429855807566e-05,
+      "loss": 3.6403,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7460043908297046e-05,
+      "loss": 3.6262,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7451657960786526e-05,
+      "loss": 3.6262,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7443272013276006e-05,
+      "loss": 3.6235,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7434886065765486e-05,
+      "loss": 3.6281,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7426500118254966e-05,
+      "loss": 3.6365,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7418114170744446e-05,
+      "loss": 3.6203,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7409744602037655e-05,
+      "loss": 3.6235,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7401375033330864e-05,
+      "loss": 3.6347,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7392989085820344e-05,
+      "loss": 3.6164,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7384603138309824e-05,
+      "loss": 3.6252,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7376217190799304e-05,
+      "loss": 3.6092,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736784762209252e-05,
+      "loss": 3.6184,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7359461674582e-05,
+      "loss": 3.6196,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735107572707148e-05,
+      "loss": 3.6119,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734268977956096e-05,
+      "loss": 3.6294,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733430383205044e-05,
+      "loss": 3.6293,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732593426334365e-05,
+      "loss": 3.6272,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731754831583313e-05,
+      "loss": 3.62,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730916236832261e-05,
+      "loss": 3.625,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730077642081209e-05,
+      "loss": 3.6278,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729239047330157e-05,
+      "loss": 3.6192,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728400452579105e-05,
+      "loss": 3.6158,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727561857828053e-05,
+      "loss": 3.6319,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726723263077001e-05,
+      "loss": 3.6167,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725886306206322e-05,
+      "loss": 3.6188,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7250477114552704e-05,
+      "loss": 3.6185,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7242091167042184e-05,
+      "loss": 3.6297,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7233705219531664e-05,
+      "loss": 3.6213,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722533565082487e-05,
+      "loss": 3.64,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721694970331435e-05,
+      "loss": 3.6146,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720858013460756e-05,
+      "loss": 3.6163,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720019418709704e-05,
+      "loss": 3.6334,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719180823958652e-05,
+      "loss": 3.6067,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7183422292076e-05,
+      "loss": 3.6183,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717503634456548e-05,
+      "loss": 3.6174,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716665039705496e-05,
+      "loss": 3.6139,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715826444954444e-05,
+      "loss": 3.6206,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714989488083766e-05,
+      "loss": 3.6242,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714150893332714e-05,
+      "loss": 3.6255,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713312298581662e-05,
+      "loss": 3.6152,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71247370383061e-05,
+      "loss": 3.6277,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711635109079557e-05,
+      "loss": 3.612,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710796514328505e-05,
+      "loss": 3.621,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709959557457827e-05,
+      "loss": 3.59,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709120962706774e-05,
+      "loss": 3.6278,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708282367955722e-05,
+      "loss": 3.6156,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70744377320467e-05,
+      "loss": 3.6249,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706605178453618e-05,
+      "loss": 3.6172,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7057682215829396e-05,
+      "loss": 3.6196,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7049296268318876e-05,
+      "loss": 3.603,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7040910320808356e-05,
+      "loss": 3.6101,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7032524373297836e-05,
+      "loss": 3.6298,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7024138425787316e-05,
+      "loss": 3.6065,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7015752478276795e-05,
+      "loss": 3.6219,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7007382909570005e-05,
+      "loss": 3.6223,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6998996962059485e-05,
+      "loss": 3.5999,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6990611014548964e-05,
+      "loss": 3.627,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6982225067038444e-05,
+      "loss": 3.6072,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6973839119527924e-05,
+      "loss": 3.6069,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6965469550821133e-05,
+      "loss": 3.6209,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6957083603310613e-05,
+      "loss": 3.6158,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6948697655800093e-05,
+      "loss": 3.6064,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694031170828957e-05,
+      "loss": 3.6176,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693192576077906e-05,
+      "loss": 3.6015,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692353981326854e-05,
+      "loss": 3.6173,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691517024456175e-05,
+      "loss": 3.6213,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690678429705123e-05,
+      "loss": 3.6172,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689839834954071e-05,
+      "loss": 3.6236,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689001240203019e-05,
+      "loss": 3.6244,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688162645451967e-05,
+      "loss": 3.6305,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687325688581288e-05,
+      "loss": 3.6097,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686487093830236e-05,
+      "loss": 3.6199,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685648499079184e-05,
+      "loss": 3.6171,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.684809904328132e-05,
+      "loss": 3.5979,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68397130957708e-05,
+      "loss": 3.6182,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6831343527064014e-05,
+      "loss": 3.6196,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6822957579553494e-05,
+      "loss": 3.6154,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6814571632042974e-05,
+      "loss": 3.6314,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6806185684532454e-05,
+      "loss": 3.6,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679779973702193e-05,
+      "loss": 3.6082,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678941378951141e-05,
+      "loss": 3.616,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678102784200089e-05,
+      "loss": 3.617,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677264189449037e-05,
+      "loss": 3.6157,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6764272325783576e-05,
+      "loss": 3.6098,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6755886378273056e-05,
+      "loss": 3.6038,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674751680956627e-05,
+      "loss": 3.6135,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673913086205575e-05,
+      "loss": 3.6115,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673074491454523e-05,
+      "loss": 3.612,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672235896703471e-05,
+      "loss": 3.6212,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671397301952419e-05,
+      "loss": 3.6199,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67056034508174e-05,
+      "loss": 3.6153,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669721750330688e-05,
+      "loss": 3.6048,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668883155579636e-05,
+      "loss": 3.6122,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668044560828584e-05,
+      "loss": 3.613,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667205966077532e-05,
+      "loss": 3.6206,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666369009206853e-05,
+      "loss": 3.6138,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665530414455801e-05,
+      "loss": 3.6095,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664691819704749e-05,
+      "loss": 3.6137,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663853224953697e-05,
+      "loss": 3.6089,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6630162680830185e-05,
+      "loss": 3.6099,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6621776733319665e-05,
+      "loss": 3.6115,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6613390785809145e-05,
+      "loss": 3.5978,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6605004838298625e-05,
+      "loss": 3.6215,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6596635269591834e-05,
+      "loss": 3.6087,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6588249322081314e-05,
+      "loss": 3.6037,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6579863374570794e-05,
+      "loss": 3.5993,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6571477427060274e-05,
+      "loss": 3.6162,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6563091479549754e-05,
+      "loss": 3.6127,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6554705532039234e-05,
+      "loss": 3.6033,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6546319584528714e-05,
+      "loss": 3.6137,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653795001582192e-05,
+      "loss": 3.6249,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65295640683114e-05,
+      "loss": 3.6185,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652117812080089e-05,
+      "loss": 3.6097,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651279217329037e-05,
+      "loss": 3.6069,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650440622577985e-05,
+      "loss": 3.6068,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649602027826933e-05,
+      "loss": 3.6173,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648763433075881e-05,
+      "loss": 3.5972,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647926476205202e-05,
+      "loss": 3.6004,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64708788145415e-05,
+      "loss": 3.6137,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646250924583471e-05,
+      "loss": 3.6088,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645412329832419e-05,
+      "loss": 3.6101,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644573735081367e-05,
+      "loss": 3.6113,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643735140330315e-05,
+      "loss": 3.6073,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642896545579263e-05,
+      "loss": 3.6072,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642057950828211e-05,
+      "loss": 3.6189,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641219356077159e-05,
+      "loss": 3.6081,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640380761326107e-05,
+      "loss": 3.6177,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639545442335801e-05,
+      "loss": 3.5986,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638706847584749e-05,
+      "loss": 3.6177,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637868252833697e-05,
+      "loss": 3.61,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637029658082645e-05,
+      "loss": 3.6169,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636191063331593e-05,
+      "loss": 3.6107,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635354106460914e-05,
+      "loss": 3.6123,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634515511709862e-05,
+      "loss": 3.6113,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63367691695881e-05,
+      "loss": 3.6261,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632838322207758e-05,
+      "loss": 3.6077,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631999727456706e-05,
+      "loss": 3.6122,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631161132705654e-05,
+      "loss": 3.6168,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630322537954602e-05,
+      "loss": 3.6022,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.62948394320355e-05,
+      "loss": 3.6079,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6286486242132446e-05,
+      "loss": 3.6148,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6278100294621926e-05,
+      "loss": 3.6095,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6269714347111406e-05,
+      "loss": 3.6106,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6261328399600886e-05,
+      "loss": 3.6116,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8456497192382812,
+      "eval_runtime": 308.4508,
+      "eval_samples_per_second": 1237.121,
+      "eval_steps_per_second": 38.661,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625294245209036e-05,
+      "loss": 3.5972,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624455650457984e-05,
+      "loss": 3.6025,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623617055706932e-05,
+      "loss": 3.6053,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.62277846095588e-05,
+      "loss": 3.61,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621939866204828e-05,
+      "loss": 3.6146,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621101271453776e-05,
+      "loss": 3.606,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6202626767027245e-05,
+      "loss": 3.6073,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6194240819516725e-05,
+      "loss": 3.6044,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6185871250809935e-05,
+      "loss": 3.6075,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6177485303299415e-05,
+      "loss": 3.6101,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6169115734592624e-05,
+      "loss": 3.6038,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6160729787082104e-05,
+      "loss": 3.5989,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615236021837531e-05,
+      "loss": 3.6138,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.614397427086479e-05,
+      "loss": 3.5958,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613558832335427e-05,
+      "loss": 3.6028,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612720237584375e-05,
+      "loss": 3.5894,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611881642833323e-05,
+      "loss": 3.6013,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611043048082271e-05,
+      "loss": 3.5997,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.61020445333122e-05,
+      "loss": 3.5898,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609365858580168e-05,
+      "loss": 3.6077,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608527263829116e-05,
+      "loss": 3.6068,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607690306958437e-05,
+      "loss": 3.6052,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606851712207385e-05,
+      "loss": 3.6013,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606013117456333e-05,
+      "loss": 3.6027,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605174522705281e-05,
+      "loss": 3.6101,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604335927954229e-05,
+      "loss": 3.6016,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.603497333203177e-05,
+      "loss": 3.5914,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602658738452125e-05,
+      "loss": 3.6125,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601821781581446e-05,
+      "loss": 3.5941,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.600983186830394e-05,
+      "loss": 3.596,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600144592079342e-05,
+      "loss": 3.6016,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59930599732829e-05,
+      "loss": 3.606,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598467402577238e-05,
+      "loss": 3.5996,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597630445706559e-05,
+      "loss": 3.6202,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596791850955507e-05,
+      "loss": 3.5949,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5959532562044546e-05,
+      "loss": 3.5979,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5951146614534026e-05,
+      "loss": 3.6139,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594277704582724e-05,
+      "loss": 3.5854,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593439109831672e-05,
+      "loss": 3.5971,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5926005150806195e-05,
+      "loss": 3.5955,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5917619203295675e-05,
+      "loss": 3.5916,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590924963458889e-05,
+      "loss": 3.6003,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590086368707837e-05,
+      "loss": 3.6025,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589247773956785e-05,
+      "loss": 3.602,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588409179205733e-05,
+      "loss": 3.5965,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5875722223350546e-05,
+      "loss": 3.6038,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.586733627584002e-05,
+      "loss": 3.5924,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58589503283295e-05,
+      "loss": 3.6033,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585056438081898e-05,
+      "loss": 3.5692,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584217843330846e-05,
+      "loss": 3.6043,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583380886460167e-05,
+      "loss": 3.5959,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582542291709115e-05,
+      "loss": 3.6062,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581703696958063e-05,
+      "loss": 3.5925,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580865102207011e-05,
+      "loss": 3.6006,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580026507455959e-05,
+      "loss": 3.5846,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5791895505852804e-05,
+      "loss": 3.588,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5783509558342284e-05,
+      "loss": 3.6113,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5775123610831764e-05,
+      "loss": 3.5839,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5766737663321244e-05,
+      "loss": 3.6008,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575836809461445e-05,
+      "loss": 3.6031,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574998214710393e-05,
+      "loss": 3.5759,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574159619959341e-05,
+      "loss": 3.6105,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573321025208289e-05,
+      "loss": 3.5875,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57248406833761e-05,
+      "loss": 3.5857,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571645473586558e-05,
+      "loss": 3.5991,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.570806878835506e-05,
+      "loss": 3.599,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569968284084454e-05,
+      "loss": 3.5854,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569129689333402e-05,
+      "loss": 3.5974,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.568291094582351e-05,
+      "loss": 3.5811,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567452499831299e-05,
+      "loss": 3.5983,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56661554296062e-05,
+      "loss": 3.6034,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565776948209568e-05,
+      "loss": 3.5978,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564938353458516e-05,
+      "loss": 3.6006,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564099758707464e-05,
+      "loss": 3.6004,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563262801836785e-05,
+      "loss": 3.6117,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562424207085733e-05,
+      "loss": 3.5908,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5615856123346807e-05,
+      "loss": 3.6037,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5607470175836287e-05,
+      "loss": 3.5897,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5599100607129496e-05,
+      "loss": 3.5828,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5590714659618976e-05,
+      "loss": 3.594,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558232871210846e-05,
+      "loss": 3.5992,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557394276459794e-05,
+      "loss": 3.5973,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556555681708742e-05,
+      "loss": 3.611,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555718724838063e-05,
+      "loss": 3.5808,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554880130087011e-05,
+      "loss": 3.5913,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554041535335959e-05,
+      "loss": 3.5943,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553202940584907e-05,
+      "loss": 3.5965,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552364345833855e-05,
+      "loss": 3.594,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.551527388963176e-05,
+      "loss": 3.5874,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550688794212124e-05,
+      "loss": 3.589,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549850199461072e-05,
+      "loss": 3.5932,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54901160471002e-05,
+      "loss": 3.5935,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5481746478393416e-05,
+      "loss": 3.5899,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5473360530882896e-05,
+      "loss": 3.5997,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5464974583372376e-05,
+      "loss": 3.5993,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5456605014665585e-05,
+      "loss": 3.5946,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5448219067155065e-05,
+      "loss": 3.5865,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5439833119644545e-05,
+      "loss": 3.5924,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5431447172134025e-05,
+      "loss": 3.5924,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54230612246235e-05,
+      "loss": 3.5989,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5414691655916714e-05,
+      "loss": 3.5923,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5406305708406194e-05,
+      "loss": 3.5882,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5397919760895674e-05,
+      "loss": 3.5942,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5389533813385154e-05,
+      "loss": 3.5917,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5381147865874634e-05,
+      "loss": 3.5922,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537277829716785e-05,
+      "loss": 3.5879,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.536439234965733e-05,
+      "loss": 3.5815,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53560064021468e-05,
+      "loss": 3.5983,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534762045463628e-05,
+      "loss": 3.5926,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53392508859295e-05,
+      "loss": 3.5822,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533086493841897e-05,
+      "loss": 3.5811,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532247899090845e-05,
+      "loss": 3.5941,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531409304339793e-05,
+      "loss": 3.5937,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530572347469115e-05,
+      "loss": 3.5836,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529733752718062e-05,
+      "loss": 3.594,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528895157967011e-05,
+      "loss": 3.6038,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528056563215959e-05,
+      "loss": 3.5945,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527217968464907e-05,
+      "loss": 3.5887,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5263810115942276e-05,
+      "loss": 3.5901,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5255424168431756e-05,
+      "loss": 3.5903,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5247038220921236e-05,
+      "loss": 3.5983,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5238652273410716e-05,
+      "loss": 3.5798,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5230282704703925e-05,
+      "loss": 3.5801,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5221896757193405e-05,
+      "loss": 3.5928,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5213510809682885e-05,
+      "loss": 3.5885,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5205124862172365e-05,
+      "loss": 3.5933,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5196755293465574e-05,
+      "loss": 3.5922,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518836934595506e-05,
+      "loss": 3.5863,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517998339844454e-05,
+      "loss": 3.5853,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517159745093402e-05,
+      "loss": 3.5983,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51632115034235e-05,
+      "loss": 3.5887,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515482555591298e-05,
+      "loss": 3.5982,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.514645598720619e-05,
+      "loss": 3.5814,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513807003969567e-05,
+      "loss": 3.5948,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512968409218515e-05,
+      "loss": 3.5939,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512129814467463e-05,
+      "loss": 3.5941,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511292857596784e-05,
+      "loss": 3.5947,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510454262845732e-05,
+      "loss": 3.5906,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50961566809468e-05,
+      "loss": 3.5919,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508777073343628e-05,
+      "loss": 3.606,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5079401164729495e-05,
+      "loss": 3.5883,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5071015217218975e-05,
+      "loss": 3.5922,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5062629269708455e-05,
+      "loss": 3.5975,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5054243322197935e-05,
+      "loss": 3.5834,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5045873753491144e-05,
+      "loss": 3.5893,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5037487805980624e-05,
+      "loss": 3.6,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5029101858470104e-05,
+      "loss": 3.5869,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5020715910959584e-05,
+      "loss": 3.5916,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501234634225279e-05,
+      "loss": 3.5916,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8445136547088623,
+      "eval_runtime": 305.0543,
+      "eval_samples_per_second": 1250.895,
+      "eval_steps_per_second": 39.091,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.500396039474227e-05,
+      "loss": 3.5845,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.499557444723175e-05,
+      "loss": 3.5788,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.498718849972123e-05,
+      "loss": 3.5842,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.497880255221071e-05,
+      "loss": 3.5934,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.497043298350393e-05,
+      "loss": 3.5888,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.496204703599341e-05,
+      "loss": 3.588,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.495366108848289e-05,
+      "loss": 3.5919,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.494527514097237e-05,
+      "loss": 3.5857,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.493690557226558e-05,
+      "loss": 3.5857,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492851962475506e-05,
+      "loss": 3.5907,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492013367724454e-05,
+      "loss": 3.5884,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491174772973402e-05,
+      "loss": 3.5772,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4903378161027226e-05,
+      "loss": 3.5966,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4894992213516706e-05,
+      "loss": 3.5793,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4886606266006186e-05,
+      "loss": 3.5821,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4878220318495666e-05,
+      "loss": 3.5718,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4869834370985146e-05,
+      "loss": 3.5819,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486144842347463e-05,
+      "loss": 3.5792,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4853062475964106e-05,
+      "loss": 3.5735,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4844676528453586e-05,
+      "loss": 3.5823,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48363069597468e-05,
+      "loss": 3.5936,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482792101223628e-05,
+      "loss": 3.5848,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4819535064725755e-05,
+      "loss": 3.5781,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4811149117215235e-05,
+      "loss": 3.5856,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.480277954850845e-05,
+      "loss": 3.5914,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.479439360099793e-05,
+      "loss": 3.5775,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4786007653487404e-05,
+      "loss": 3.5741,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4777621705976884e-05,
+      "loss": 3.592,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476923575846637e-05,
+      "loss": 3.5788,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476086618975958e-05,
+      "loss": 3.5749,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.475248024224906e-05,
+      "loss": 3.5824,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.474409429473854e-05,
+      "loss": 3.5848,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.473570834722802e-05,
+      "loss": 3.5856,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.472733877852123e-05,
+      "loss": 3.5994,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471895283101071e-05,
+      "loss": 3.5718,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471056688350019e-05,
+      "loss": 3.5854,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470218093598967e-05,
+      "loss": 3.5934,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.469381136728288e-05,
+      "loss": 3.5628,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468542541977236e-05,
+      "loss": 3.581,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467703947226184e-05,
+      "loss": 3.579,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4668653524751324e-05,
+      "loss": 3.5693,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4660283956044533e-05,
+      "loss": 3.5888,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465189800853401e-05,
+      "loss": 3.5792,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464351206102349e-05,
+      "loss": 3.5863,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463512611351297e-05,
+      "loss": 3.5759,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462675654480618e-05,
+      "loss": 3.5847,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461837059729566e-05,
+      "loss": 3.5717,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460998464978514e-05,
+      "loss": 3.5858,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460159870227462e-05,
+      "loss": 3.5543,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45932127547641e-05,
+      "loss": 3.5841,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.458484318605731e-05,
+      "loss": 3.5784,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.457645723854679e-05,
+      "loss": 3.5889,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.456807129103628e-05,
+      "loss": 3.5719,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455968534352576e-05,
+      "loss": 3.5774,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455131577481897e-05,
+      "loss": 3.5704,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.454292982730845e-05,
+      "loss": 3.5715,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.453454387979793e-05,
+      "loss": 3.5939,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.452615793228741e-05,
+      "loss": 3.5635,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.451777198477689e-05,
+      "loss": 3.5867,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4509402416070096e-05,
+      "loss": 3.5832,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4501016468559576e-05,
+      "loss": 3.5594,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4492630521049056e-05,
+      "loss": 3.5893,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4484244573538536e-05,
+      "loss": 3.5676,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4475875004831745e-05,
+      "loss": 3.5677,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.446748905732123e-05,
+      "loss": 3.5792,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445910310981071e-05,
+      "loss": 3.5818,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445071716230019e-05,
+      "loss": 3.5661,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44423475935934e-05,
+      "loss": 3.5792,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443396164608288e-05,
+      "loss": 3.5628,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442557569857236e-05,
+      "loss": 3.5769,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.441718975106184e-05,
+      "loss": 3.5872,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440882018235505e-05,
+      "loss": 3.5773,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440043423484453e-05,
+      "loss": 3.5834,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439204828733401e-05,
+      "loss": 3.5825,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.438366233982349e-05,
+      "loss": 3.5926,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.43752927711167e-05,
+      "loss": 3.5746,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4366906823606185e-05,
+      "loss": 3.585,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4358520876095665e-05,
+      "loss": 3.5727,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4350134928585145e-05,
+      "loss": 3.5626,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4341765359878354e-05,
+      "loss": 3.5765,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4333379412367834e-05,
+      "loss": 3.5806,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4324993464857314e-05,
+      "loss": 3.5817,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4316607517346794e-05,
+      "loss": 3.5906,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4308221569836274e-05,
+      "loss": 3.5678,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4299835622325754e-05,
+      "loss": 3.5677,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429146605361896e-05,
+      "loss": 3.573,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.428308010610844e-05,
+      "loss": 3.5846,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.427469415859792e-05,
+      "loss": 3.5729,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.42663082110874e-05,
+      "loss": 3.5703,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.425793864238062e-05,
+      "loss": 3.5722,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.42495526948701e-05,
+      "loss": 3.5733,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424116674735958e-05,
+      "loss": 3.5731,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.423278079984906e-05,
+      "loss": 3.5723,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422439485233854e-05,
+      "loss": 3.5807,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.421602528363175e-05,
+      "loss": 3.5805,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.420763933612123e-05,
+      "loss": 3.5785,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.419925338861071e-05,
+      "loss": 3.5666,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.419086744110018e-05,
+      "loss": 3.5724,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41824978723934e-05,
+      "loss": 3.5737,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.417411192488288e-05,
+      "loss": 3.5874,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.416572597737236e-05,
+      "loss": 3.5704,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.415734002986184e-05,
+      "loss": 3.5697,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414895408235132e-05,
+      "loss": 3.5766,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414058451364453e-05,
+      "loss": 3.5722,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413219856613401e-05,
+      "loss": 3.5766,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4123812618623486e-05,
+      "loss": 3.5724,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4115426671112966e-05,
+      "loss": 3.5601,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4107040723602446e-05,
+      "loss": 3.5819,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4098671154895655e-05,
+      "loss": 3.5713,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4090285207385135e-05,
+      "loss": 3.5686,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4081899259874615e-05,
+      "loss": 3.5609,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4073513312364095e-05,
+      "loss": 3.5763,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.406514374365731e-05,
+      "loss": 3.5786,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.405675779614679e-05,
+      "loss": 3.562,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.404837184863627e-05,
+      "loss": 3.5771,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403998590112575e-05,
+      "loss": 3.5827,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403159995361523e-05,
+      "loss": 3.583,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.402323038490844e-05,
+      "loss": 3.5701,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.401484443739792e-05,
+      "loss": 3.571,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40064584898874e-05,
+      "loss": 3.5701,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.399807254237688e-05,
+      "loss": 3.5824,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398970297367009e-05,
+      "loss": 3.5588,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398131702615957e-05,
+      "loss": 3.5667,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.397293107864905e-05,
+      "loss": 3.5676,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396454513113853e-05,
+      "loss": 3.5738,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3956175562431744e-05,
+      "loss": 3.5755,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3947789614921224e-05,
+      "loss": 3.5746,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3939403667410704e-05,
+      "loss": 3.5648,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3931017719900184e-05,
+      "loss": 3.5671,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392264815119339e-05,
+      "loss": 3.5825,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391426220368287e-05,
+      "loss": 3.5714,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390587625617235e-05,
+      "loss": 3.5825,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389749030866183e-05,
+      "loss": 3.56,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388912073995504e-05,
+      "loss": 3.578,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388073479244452e-05,
+      "loss": 3.5755,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3872348844934e-05,
+      "loss": 3.5752,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.386396289742348e-05,
+      "loss": 3.5781,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.385557694991296e-05,
+      "loss": 3.5707,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.384720738120618e-05,
+      "loss": 3.5776,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.383882143369566e-05,
+      "loss": 3.5867,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.383043548618514e-05,
+      "loss": 3.5734,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382204953867462e-05,
+      "loss": 3.5728,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.381367996996783e-05,
+      "loss": 3.5788,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3805294022457307e-05,
+      "loss": 3.5633,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3796908074946787e-05,
+      "loss": 3.5698,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3788522127436266e-05,
+      "loss": 3.5827,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3780136179925746e-05,
+      "loss": 3.5707,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3771766611218956e-05,
+      "loss": 3.5747,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3763380663708435e-05,
+      "loss": 3.5708,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8452529907226562,
+      "eval_runtime": 322.3677,
+      "eval_samples_per_second": 1183.714,
+      "eval_steps_per_second": 36.992,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3754994716197915e-05,
+      "loss": 3.5785,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.37466087686874e-05,
+      "loss": 3.5607,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373822282117688e-05,
+      "loss": 3.5682,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372983687366636e-05,
+      "loss": 3.5768,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372146730495957e-05,
+      "loss": 3.5747,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371308135744905e-05,
+      "loss": 3.5722,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370469540993853e-05,
+      "loss": 3.5663,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369630946242801e-05,
+      "loss": 3.5697,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368792351491749e-05,
+      "loss": 3.5686,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36795539462107e-05,
+      "loss": 3.5781,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367116799870018e-05,
+      "loss": 3.5646,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366278205118966e-05,
+      "loss": 3.5629,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365441248248287e-05,
+      "loss": 3.5809,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3646026534972356e-05,
+      "loss": 3.5578,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3637640587461836e-05,
+      "loss": 3.5656,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3629254639951316e-05,
+      "loss": 3.5544,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362086869244079e-05,
+      "loss": 3.5652,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361248274493027e-05,
+      "loss": 3.5599,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.360409679741975e-05,
+      "loss": 3.5582,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.359571084990923e-05,
+      "loss": 3.5604,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.358734128120244e-05,
+      "loss": 3.578,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357895533369192e-05,
+      "loss": 3.5668,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35705693861814e-05,
+      "loss": 3.5634,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356218343867088e-05,
+      "loss": 3.5682,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3553813869964094e-05,
+      "loss": 3.5781,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3545427922453574e-05,
+      "loss": 3.5589,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3537041974943054e-05,
+      "loss": 3.558,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3528656027432534e-05,
+      "loss": 3.5746,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3520270079922013e-05,
+      "loss": 3.5582,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351190051121522e-05,
+      "loss": 3.5609,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.35035145637047e-05,
+      "loss": 3.5635,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349512861619418e-05,
+      "loss": 3.5641,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348674266868366e-05,
+      "loss": 3.5733,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347837309997687e-05,
+      "loss": 3.5771,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346998715246635e-05,
+      "loss": 3.56,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346160120495583e-05,
+      "loss": 3.564,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345321525744531e-05,
+      "loss": 3.5746,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344482930993479e-05,
+      "loss": 3.5485,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343645974122801e-05,
+      "loss": 3.5682,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342807379371749e-05,
+      "loss": 3.5576,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341968784620697e-05,
+      "loss": 3.5537,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341130189869645e-05,
+      "loss": 3.5669,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3402932329989656e-05,
+      "loss": 3.5708,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3394546382479136e-05,
+      "loss": 3.5652,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3386160434968616e-05,
+      "loss": 3.5581,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3377790866261825e-05,
+      "loss": 3.5644,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3369404918751305e-05,
+      "loss": 3.5593,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3361018971240785e-05,
+      "loss": 3.5648,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3352633023730265e-05,
+      "loss": 3.5403,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3344247076219745e-05,
+      "loss": 3.5651,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333587750751296e-05,
+      "loss": 3.5637,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332749156000244e-05,
+      "loss": 3.5712,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331910561249192e-05,
+      "loss": 3.5568,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33107196649814e-05,
+      "loss": 3.5606,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330235009627461e-05,
+      "loss": 3.5517,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329396414876409e-05,
+      "loss": 3.552,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.328557820125357e-05,
+      "loss": 3.5759,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.327719225374305e-05,
+      "loss": 3.542,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326882268503626e-05,
+      "loss": 3.5704,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326043673752574e-05,
+      "loss": 3.566,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325205079001522e-05,
+      "loss": 3.5464,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.32436648425047e-05,
+      "loss": 3.5717,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323527889499418e-05,
+      "loss": 3.5489,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3226909326287395e-05,
+      "loss": 3.5487,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3218523378776874e-05,
+      "loss": 3.5609,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3210137431266354e-05,
+      "loss": 3.5666,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3201751483755834e-05,
+      "loss": 3.5452,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3193381915049043e-05,
+      "loss": 3.5683,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3184995967538523e-05,
+      "loss": 3.5436,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3176610020028e-05,
+      "loss": 3.5574,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316822407251748e-05,
+      "loss": 3.5711,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315985450381069e-05,
+      "loss": 3.5616,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315146855630017e-05,
+      "loss": 3.5646,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314308260878965e-05,
+      "loss": 3.5664,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.313469666127913e-05,
+      "loss": 3.577,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.312632709257235e-05,
+      "loss": 3.5555,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.311794114506183e-05,
+      "loss": 3.5677,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310955519755131e-05,
+      "loss": 3.5501,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310116925004079e-05,
+      "loss": 3.5472,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309278330253027e-05,
+      "loss": 3.5605,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308441373382348e-05,
+      "loss": 3.5596,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307602778631296e-05,
+      "loss": 3.5665,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306764183880244e-05,
+      "loss": 3.575,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305925589129192e-05,
+      "loss": 3.5575,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3050886322585126e-05,
+      "loss": 3.547,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3042500375074606e-05,
+      "loss": 3.5529,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3034114427564086e-05,
+      "loss": 3.5683,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302572848005357e-05,
+      "loss": 3.555,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301735891134678e-05,
+      "loss": 3.5528,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300897296383626e-05,
+      "loss": 3.5569,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.300058701632574e-05,
+      "loss": 3.5523,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299220106881522e-05,
+      "loss": 3.5575,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298383150010843e-05,
+      "loss": 3.5561,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297544555259791e-05,
+      "loss": 3.5646,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296705960508739e-05,
+      "loss": 3.5675,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2958673657576864e-05,
+      "loss": 3.5571,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295030408887008e-05,
+      "loss": 3.5528,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294191814135956e-05,
+      "loss": 3.5542,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293353219384904e-05,
+      "loss": 3.5607,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292514624633852e-05,
+      "loss": 3.567,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2916776677631735e-05,
+      "loss": 3.557,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2908390730121215e-05,
+      "loss": 3.5515,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2900004782610695e-05,
+      "loss": 3.562,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289161883510017e-05,
+      "loss": 3.5544,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2883249266393384e-05,
+      "loss": 3.5563,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2874863318882864e-05,
+      "loss": 3.5563,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286647737137234e-05,
+      "loss": 3.5459,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285809142386182e-05,
+      "loss": 3.5623,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2849721855155033e-05,
+      "loss": 3.5568,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284133590764451e-05,
+      "loss": 3.548,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283294996013399e-05,
+      "loss": 3.5455,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282456401262347e-05,
+      "loss": 3.5579,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281617806511295e-05,
+      "loss": 3.5626,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280780849640617e-05,
+      "loss": 3.5499,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279942254889564e-05,
+      "loss": 3.5577,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279103660138512e-05,
+      "loss": 3.5618,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.27826506538746e-05,
+      "loss": 3.5685,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.277428108516781e-05,
+      "loss": 3.5578,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276589513765729e-05,
+      "loss": 3.5507,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275750919014677e-05,
+      "loss": 3.5536,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274912324263625e-05,
+      "loss": 3.5646,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274073729512573e-05,
+      "loss": 3.5425,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273236772641895e-05,
+      "loss": 3.5509,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.272398177890843e-05,
+      "loss": 3.5519,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.271559583139791e-05,
+      "loss": 3.5539,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270720988388739e-05,
+      "loss": 3.5566,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2698840315180596e-05,
+      "loss": 3.5651,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2690454367670076e-05,
+      "loss": 3.5452,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2682068420159556e-05,
+      "loss": 3.552,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2673682472649036e-05,
+      "loss": 3.566,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2665312903942245e-05,
+      "loss": 3.5564,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2656926956431725e-05,
+      "loss": 3.5644,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2648541008921205e-05,
+      "loss": 3.543,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2640155061410685e-05,
+      "loss": 3.5604,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26317854927039e-05,
+      "loss": 3.5597,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262339954519338e-05,
+      "loss": 3.5565,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261501359768286e-05,
+      "loss": 3.5564,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260662765017234e-05,
+      "loss": 3.5545,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259825808146555e-05,
+      "loss": 3.5642,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258987213395503e-05,
+      "loss": 3.5628,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258148618644451e-05,
+      "loss": 3.5624,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257310023893399e-05,
+      "loss": 3.5537,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.25647306702272e-05,
+      "loss": 3.5638,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255634472271668e-05,
+      "loss": 3.552,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254795877520616e-05,
+      "loss": 3.549,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253957282769564e-05,
+      "loss": 3.5691,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2531203258988854e-05,
+      "loss": 3.5511,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2522817311478334e-05,
+      "loss": 3.5628,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2514431363967814e-05,
+      "loss": 3.5505,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8457460403442383,
+      "eval_runtime": 304.8763,
+      "eval_samples_per_second": 1251.626,
+      "eval_steps_per_second": 39.114,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2506045416457294e-05,
+      "loss": 3.5476,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24976758477505e-05,
+      "loss": 3.548,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248928990023998e-05,
+      "loss": 3.551,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248090395272946e-05,
+      "loss": 3.5574,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247251800521894e-05,
+      "loss": 3.559,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246414843651215e-05,
+      "loss": 3.5544,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.245576248900163e-05,
+      "loss": 3.5479,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.244737654149111e-05,
+      "loss": 3.551,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243899059398059e-05,
+      "loss": 3.5519,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243062102527381e-05,
+      "loss": 3.5635,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242223507776329e-05,
+      "loss": 3.5483,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.241384913025277e-05,
+      "loss": 3.5501,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.240547956154598e-05,
+      "loss": 3.5643,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239709361403546e-05,
+      "loss": 3.5412,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238870766652494e-05,
+      "loss": 3.5468,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238032171901442e-05,
+      "loss": 3.5412,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23719357715039e-05,
+      "loss": 3.5503,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236354982399338e-05,
+      "loss": 3.5378,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235516387648286e-05,
+      "loss": 3.5412,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.234677792897234e-05,
+      "loss": 3.5493,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233839198146182e-05,
+      "loss": 3.5573,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2330022412755026e-05,
+      "loss": 3.5529,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232163646524451e-05,
+      "loss": 3.5436,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.231325051773399e-05,
+      "loss": 3.5553,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23048809490272e-05,
+      "loss": 3.56,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.229649500151668e-05,
+      "loss": 3.5412,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228810905400616e-05,
+      "loss": 3.541,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227972310649564e-05,
+      "loss": 3.5593,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227133715898512e-05,
+      "loss": 3.5385,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226296759027833e-05,
+      "loss": 3.5454,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.225458164276781e-05,
+      "loss": 3.5441,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.224619569525729e-05,
+      "loss": 3.5495,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.223780974774677e-05,
+      "loss": 3.5545,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222944017903998e-05,
+      "loss": 3.5608,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2221054231529466e-05,
+      "loss": 3.5487,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2212668284018946e-05,
+      "loss": 3.5476,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.220428233650842e-05,
+      "loss": 3.5571,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2195912767801635e-05,
+      "loss": 3.5339,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2187526820291115e-05,
+      "loss": 3.5536,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2179140872780595e-05,
+      "loss": 3.5387,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217075492527007e-05,
+      "loss": 3.5385,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2162385356563284e-05,
+      "loss": 3.5463,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2153999409052764e-05,
+      "loss": 3.5561,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2145613461542244e-05,
+      "loss": 3.5454,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.213722751403172e-05,
+      "loss": 3.5446,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2128841566521204e-05,
+      "loss": 3.5481,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.212047199781442e-05,
+      "loss": 3.5438,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211208605030389e-05,
+      "loss": 3.5501,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210370010279337e-05,
+      "loss": 3.5234,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209531415528285e-05,
+      "loss": 3.5469,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.208694458657607e-05,
+      "loss": 3.5485,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207855863906554e-05,
+      "loss": 3.5554,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207017269155502e-05,
+      "loss": 3.5413,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20617867440445e-05,
+      "loss": 3.5407,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205341717533772e-05,
+      "loss": 3.5381,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.204503122782719e-05,
+      "loss": 3.5346,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203664528031667e-05,
+      "loss": 3.5602,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202825933280616e-05,
+      "loss": 3.5318,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201988976409937e-05,
+      "loss": 3.5506,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201150381658885e-05,
+      "loss": 3.549,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2003117869078327e-05,
+      "loss": 3.5332,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1994731921567807e-05,
+      "loss": 3.5544,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1986362352861016e-05,
+      "loss": 3.5317,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1977976405350496e-05,
+      "loss": 3.5351,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1969590457839976e-05,
+      "loss": 3.5424,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1961204510329456e-05,
+      "loss": 3.5531,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1952834941622665e-05,
+      "loss": 3.5293,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1944448994112145e-05,
+      "loss": 3.5504,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1936063046601625e-05,
+      "loss": 3.5257,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1927677099091104e-05,
+      "loss": 3.5426,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191930753038432e-05,
+      "loss": 3.5542,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19109215828738e-05,
+      "loss": 3.5468,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190253563536328e-05,
+      "loss": 3.546,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189414968785276e-05,
+      "loss": 3.5515,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188578011914597e-05,
+      "loss": 3.5582,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187739417163545e-05,
+      "loss": 3.5431,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186900822412493e-05,
+      "loss": 3.5543,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186062227661441e-05,
+      "loss": 3.5353,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185225270790762e-05,
+      "loss": 3.5311,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18438667603971e-05,
+      "loss": 3.5426,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.183548081288658e-05,
+      "loss": 3.5444,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182709486537606e-05,
+      "loss": 3.5463,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1818708917865545e-05,
+      "loss": 3.5608,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1810339349158754e-05,
+      "loss": 3.5428,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1801953401648234e-05,
+      "loss": 3.5295,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1793567454137714e-05,
+      "loss": 3.5414,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1785181506627194e-05,
+      "loss": 3.5489,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1776795559116674e-05,
+      "loss": 3.5405,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176842599040988e-05,
+      "loss": 3.5376,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176004004289936e-05,
+      "loss": 3.5405,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.175165409538884e-05,
+      "loss": 3.5352,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174326814787832e-05,
+      "loss": 3.5447,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173489857917153e-05,
+      "loss": 3.5364,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172651263166101e-05,
+      "loss": 3.5477,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.17181266841505e-05,
+      "loss": 3.554,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170974073663998e-05,
+      "loss": 3.5434,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170137116793319e-05,
+      "loss": 3.5335,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169298522042267e-05,
+      "loss": 3.5402,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168459927291215e-05,
+      "loss": 3.5402,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.167621332540163e-05,
+      "loss": 3.5504,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1667843756694837e-05,
+      "loss": 3.5423,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1659457809184317e-05,
+      "loss": 3.5375,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1651071861673796e-05,
+      "loss": 3.5418,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1642685914163276e-05,
+      "loss": 3.5396,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1634316345456486e-05,
+      "loss": 3.5435,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1625930397945965e-05,
+      "loss": 3.5401,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161754445043545e-05,
+      "loss": 3.5327,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160915850292493e-05,
+      "loss": 3.5446,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160078893421814e-05,
+      "loss": 3.5406,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159240298670762e-05,
+      "loss": 3.5305,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15840170391971e-05,
+      "loss": 3.5312,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157563109168658e-05,
+      "loss": 3.5356,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156726152297979e-05,
+      "loss": 3.5515,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155887557546927e-05,
+      "loss": 3.5355,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155048962795875e-05,
+      "loss": 3.5412,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154210368044823e-05,
+      "loss": 3.54,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153373411174144e-05,
+      "loss": 3.5598,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152534816423092e-05,
+      "loss": 3.5418,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1516962216720406e-05,
+      "loss": 3.5354,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1508576269209886e-05,
+      "loss": 3.5409,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1500206700503095e-05,
+      "loss": 3.5439,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1491820752992575e-05,
+      "loss": 3.5279,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1483434805482055e-05,
+      "loss": 3.5343,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1475048857971535e-05,
+      "loss": 3.5355,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1466679289264744e-05,
+      "loss": 3.5405,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1458293341754224e-05,
+      "loss": 3.5369,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1449907394243704e-05,
+      "loss": 3.5476,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1441521446733184e-05,
+      "loss": 3.5365,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.143315187802639e-05,
+      "loss": 3.5369,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.142476593051587e-05,
+      "loss": 3.5486,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.141637998300536e-05,
+      "loss": 3.542,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140799403549484e-05,
+      "loss": 3.5441,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139962446678805e-05,
+      "loss": 3.5314,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139123851927753e-05,
+      "loss": 3.5415,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138285257176701e-05,
+      "loss": 3.547,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.137446662425649e-05,
+      "loss": 3.5375,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13660970555497e-05,
+      "loss": 3.5448,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135771110803918e-05,
+      "loss": 3.538,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134932516052866e-05,
+      "loss": 3.5506,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134093921301814e-05,
+      "loss": 3.5485,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1332569644311347e-05,
+      "loss": 3.546,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1324183696800826e-05,
+      "loss": 3.5399,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.131579774929031e-05,
+      "loss": 3.5445,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.130741180177979e-05,
+      "loss": 3.533,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1299042233073e-05,
+      "loss": 3.5334,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129065628556248e-05,
+      "loss": 3.5537,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128227033805196e-05,
+      "loss": 3.5343,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127388439054144e-05,
+      "loss": 3.5467,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126551482183465e-05,
+      "loss": 3.5405,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.846527576446533,
+      "eval_runtime": 304.4562,
+      "eval_samples_per_second": 1253.353,
+      "eval_steps_per_second": 39.168,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125712887432413e-05,
+      "loss": 3.5238,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124874292681361e-05,
+      "loss": 3.5317,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124035697930309e-05,
+      "loss": 3.5353,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1231971031792564e-05,
+      "loss": 3.5413,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122360146308578e-05,
+      "loss": 3.5412,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121521551557526e-05,
+      "loss": 3.5434,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120682956806475e-05,
+      "loss": 3.5339,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119844362055422e-05,
+      "loss": 3.5347,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1190074051847436e-05,
+      "loss": 3.5389,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1181688104336916e-05,
+      "loss": 3.5496,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117330215682639e-05,
+      "loss": 3.528,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116491620931587e-05,
+      "loss": 3.536,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1156546640609085e-05,
+      "loss": 3.5437,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1148160693098565e-05,
+      "loss": 3.5317,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113977474558804e-05,
+      "loss": 3.5313,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113138879807752e-05,
+      "loss": 3.528,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1123019229370734e-05,
+      "loss": 3.5315,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1114633281860214e-05,
+      "loss": 3.5291,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1106247334349694e-05,
+      "loss": 3.5275,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1097861386839174e-05,
+      "loss": 3.5291,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1089475439328654e-05,
+      "loss": 3.5425,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108110587062186e-05,
+      "loss": 3.5336,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107271992311134e-05,
+      "loss": 3.5315,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.106433397560082e-05,
+      "loss": 3.5371,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.10559480280903e-05,
+      "loss": 3.548,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.104757845938351e-05,
+      "loss": 3.5235,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103919251187299e-05,
+      "loss": 3.5283,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103080656436247e-05,
+      "loss": 3.539,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102242061685195e-05,
+      "loss": 3.5274,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101405104814517e-05,
+      "loss": 3.5296,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100566510063465e-05,
+      "loss": 3.5275,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099727915312413e-05,
+      "loss": 3.5364,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098889320561361e-05,
+      "loss": 3.5392,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0980523636906816e-05,
+      "loss": 3.544,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0972137689396296e-05,
+      "loss": 3.537,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0963751741885776e-05,
+      "loss": 3.5316,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0955365794375256e-05,
+      "loss": 3.5404,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0946996225668465e-05,
+      "loss": 3.5178,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0938610278157945e-05,
+      "loss": 3.5405,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0930224330647425e-05,
+      "loss": 3.5201,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0921838383136905e-05,
+      "loss": 3.5273,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091346881443012e-05,
+      "loss": 3.5303,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09050828669196e-05,
+      "loss": 3.5375,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089669691940908e-05,
+      "loss": 3.5331,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088831097189856e-05,
+      "loss": 3.5293,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087994140319177e-05,
+      "loss": 3.5324,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087155545568125e-05,
+      "loss": 3.5292,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086316950817073e-05,
+      "loss": 3.5367,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085478356066021e-05,
+      "loss": 3.5041,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084641399195342e-05,
+      "loss": 3.5286,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08380280444429e-05,
+      "loss": 3.5391,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082964209693238e-05,
+      "loss": 3.537,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082125614942186e-05,
+      "loss": 3.5223,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0812870201911346e-05,
+      "loss": 3.5271,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0804500633204555e-05,
+      "loss": 3.5246,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0796114685694035e-05,
+      "loss": 3.5179,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0787728738183515e-05,
+      "loss": 3.5504,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0779342790672995e-05,
+      "loss": 3.5144,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0770973221966204e-05,
+      "loss": 3.5363,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0762587274455684e-05,
+      "loss": 3.5279,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0754201326945164e-05,
+      "loss": 3.5218,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0745815379434644e-05,
+      "loss": 3.5341,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073744581072785e-05,
+      "loss": 3.5175,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072905986321733e-05,
+      "loss": 3.5224,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072067391570681e-05,
+      "loss": 3.5216,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07122879681963e-05,
+      "loss": 3.5421,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070391839948951e-05,
+      "loss": 3.5126,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069553245197899e-05,
+      "loss": 3.5345,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.068714650446847e-05,
+      "loss": 3.5123,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067877693576168e-05,
+      "loss": 3.5237,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067039098825116e-05,
+      "loss": 3.5405,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066200504074064e-05,
+      "loss": 3.5362,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.065361909323012e-05,
+      "loss": 3.5282,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0645249524523326e-05,
+      "loss": 3.5342,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0636863577012806e-05,
+      "loss": 3.5431,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0628477629502286e-05,
+      "loss": 3.5265,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0620091681991766e-05,
+      "loss": 3.5394,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061172211328498e-05,
+      "loss": 3.5222,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060333616577446e-05,
+      "loss": 3.5182,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059495021826394e-05,
+      "loss": 3.5261,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058656427075342e-05,
+      "loss": 3.529,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057819470204663e-05,
+      "loss": 3.5333,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056980875453611e-05,
+      "loss": 3.5423,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056142280702559e-05,
+      "loss": 3.5326,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055303685951507e-05,
+      "loss": 3.513,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054466729080828e-05,
+      "loss": 3.5233,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053628134329776e-05,
+      "loss": 3.5353,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052789539578724e-05,
+      "loss": 3.526,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051950944827672e-05,
+      "loss": 3.522,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0511123500766203e-05,
+      "loss": 3.526,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0502753932059412e-05,
+      "loss": 3.5217,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0494367984548892e-05,
+      "loss": 3.5275,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0485982037038376e-05,
+      "loss": 3.5239,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0477596089527856e-05,
+      "loss": 3.5321,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0469226520821065e-05,
+      "loss": 3.5352,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0460840573310545e-05,
+      "loss": 3.5307,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0452454625800025e-05,
+      "loss": 3.5192,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0444068678289505e-05,
+      "loss": 3.5275,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0435699109582717e-05,
+      "loss": 3.527,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0427313162072197e-05,
+      "loss": 3.5349,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0418927214561677e-05,
+      "loss": 3.529,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0410557645854886e-05,
+      "loss": 3.5202,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0402171698344366e-05,
+      "loss": 3.5274,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0393785750833846e-05,
+      "loss": 3.5278,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.038539980332333e-05,
+      "loss": 3.529,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.037701385581281e-05,
+      "loss": 3.5255,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036864428710602e-05,
+      "loss": 3.5168,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.03602583395955e-05,
+      "loss": 3.5312,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0351872392084978e-05,
+      "loss": 3.5284,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0343486444574458e-05,
+      "loss": 3.5129,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033511687586767e-05,
+      "loss": 3.5187,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032673092835715e-05,
+      "loss": 3.5192,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.031834498084663e-05,
+      "loss": 3.5352,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.030995903333611e-05,
+      "loss": 3.5212,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.030158946462932e-05,
+      "loss": 3.5242,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.02932035171188e-05,
+      "loss": 3.5301,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0284817569608283e-05,
+      "loss": 3.5439,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0276431622097763e-05,
+      "loss": 3.5249,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0268062053390972e-05,
+      "loss": 3.5185,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0259676105880452e-05,
+      "loss": 3.5284,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0251290158369932e-05,
+      "loss": 3.5257,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0242904210859412e-05,
+      "loss": 3.5156,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0234534642152624e-05,
+      "loss": 3.5178,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0226148694642104e-05,
+      "loss": 3.5229,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0217762747131584e-05,
+      "loss": 3.5256,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0209376799621064e-05,
+      "loss": 3.5228,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0201007230914273e-05,
+      "loss": 3.5343,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0192621283403753e-05,
+      "loss": 3.5187,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0184235335893237e-05,
+      "loss": 3.5246,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.017584938838271e-05,
+      "loss": 3.5255,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0167479819675926e-05,
+      "loss": 3.5351,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0159093872165406e-05,
+      "loss": 3.5312,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0150707924654886e-05,
+      "loss": 3.5144,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0142321977144362e-05,
+      "loss": 3.5264,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0133952408437578e-05,
+      "loss": 3.5331,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0125566460927058e-05,
+      "loss": 3.5232,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0117180513416538e-05,
+      "loss": 3.5302,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010879456590601e-05,
+      "loss": 3.523,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0100424997199227e-05,
+      "loss": 3.5381,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0092039049688707e-05,
+      "loss": 3.5285,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0083653102178184e-05,
+      "loss": 3.5317,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0075267154667663e-05,
+      "loss": 3.5227,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.006689758596088e-05,
+      "loss": 3.5343,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.005851163845036e-05,
+      "loss": 3.5189,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0050125690939836e-05,
+      "loss": 3.5148,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0041739743429316e-05,
+      "loss": 3.5404,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0033370174722532e-05,
+      "loss": 3.525,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.002498422721201e-05,
+      "loss": 3.5289,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0016598279701485e-05,
+      "loss": 3.5263,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0008212332190965e-05,
+      "loss": 3.5146,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8489813804626465,
+      "eval_runtime": 302.425,
+      "eval_samples_per_second": 1261.771,
+      "eval_steps_per_second": 39.431,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.999984276348418e-05,
+      "loss": 3.5171,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9991456815973657e-05,
+      "loss": 3.5202,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9983070868463137e-05,
+      "loss": 3.5274,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9974684920952617e-05,
+      "loss": 3.5233,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9966315352245833e-05,
+      "loss": 3.5308,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9957929404735306e-05,
+      "loss": 3.5189,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.994954345722479e-05,
+      "loss": 3.5214,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.994115750971427e-05,
+      "loss": 3.5195,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9932787941007485e-05,
+      "loss": 3.5372,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.992440199349696e-05,
+      "loss": 3.5146,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.991601604598644e-05,
+      "loss": 3.5228,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9907646477279654e-05,
+      "loss": 3.5291,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989926052976913e-05,
+      "loss": 3.5205,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989087458225861e-05,
+      "loss": 3.5154,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.988248863474809e-05,
+      "loss": 3.5116,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.987410268723757e-05,
+      "loss": 3.5119,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.986571673972705e-05,
+      "loss": 3.5142,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.985733079221653e-05,
+      "loss": 3.5159,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.984894484470601e-05,
+      "loss": 3.5117,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9840575275999223e-05,
+      "loss": 3.5307,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9832189328488703e-05,
+      "loss": 3.5203,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9823803380978183e-05,
+      "loss": 3.5138,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9815417433467663e-05,
+      "loss": 3.5222,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9807047864760872e-05,
+      "loss": 3.5331,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9798661917250352e-05,
+      "loss": 3.5101,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9790275969739832e-05,
+      "loss": 3.5158,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9781890022229315e-05,
+      "loss": 3.5194,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9773520453522524e-05,
+      "loss": 3.5199,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9765134506012004e-05,
+      "loss": 3.5109,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9756748558501484e-05,
+      "loss": 3.5117,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9748362610990964e-05,
+      "loss": 3.525,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9739993042284177e-05,
+      "loss": 3.5233,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9731607094773657e-05,
+      "loss": 3.5263,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9723221147263137e-05,
+      "loss": 3.5223,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9714835199752617e-05,
+      "loss": 3.5172,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9706449252242097e-05,
+      "loss": 3.5219,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9698079683535306e-05,
+      "loss": 3.5068,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9689693736024786e-05,
+      "loss": 3.5252,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.968130778851427e-05,
+      "loss": 3.5106,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967292184100375e-05,
+      "loss": 3.5114,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9664552272296958e-05,
+      "loss": 3.5131,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9656166324786438e-05,
+      "loss": 3.5303,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9647780377275918e-05,
+      "loss": 3.5176,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9639394429765398e-05,
+      "loss": 3.5143,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.963102486105861e-05,
+      "loss": 3.5223,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962263891354809e-05,
+      "loss": 3.5118,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.961425296603757e-05,
+      "loss": 3.522,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.960586701852705e-05,
+      "loss": 3.4959,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.959749744982026e-05,
+      "loss": 3.5105,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.958911150230974e-05,
+      "loss": 3.5226,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9580725554799223e-05,
+      "loss": 3.5244,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9572339607288703e-05,
+      "loss": 3.5118,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9563970038581912e-05,
+      "loss": 3.5113,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9555584091071392e-05,
+      "loss": 3.5131,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.954719814356087e-05,
+      "loss": 3.4994,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.953881219605035e-05,
+      "loss": 3.5328,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9530442627343564e-05,
+      "loss": 3.5073,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9522056679833044e-05,
+      "loss": 3.5198,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9513670732322524e-05,
+      "loss": 3.5165,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9505284784812004e-05,
+      "loss": 3.504,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9496915216105213e-05,
+      "loss": 3.518,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9488529268594693e-05,
+      "loss": 3.508,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9480143321084176e-05,
+      "loss": 3.5097,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9471773752377385e-05,
+      "loss": 3.5037,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9463387804866865e-05,
+      "loss": 3.5294,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9455001857356345e-05,
+      "loss": 3.5021,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9446615909845825e-05,
+      "loss": 3.5129,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9438246341139038e-05,
+      "loss": 3.4984,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9429860393628518e-05,
+      "loss": 3.5151,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9421474446117998e-05,
+      "loss": 3.5217,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9413088498607478e-05,
+      "loss": 3.5224,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9404718929900687e-05,
+      "loss": 3.5138,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9396332982390167e-05,
+      "loss": 3.5262,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9387947034879647e-05,
+      "loss": 3.5287,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.937956108736913e-05,
+      "loss": 3.5127,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.937119151866234e-05,
+      "loss": 3.5235,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.936280557115182e-05,
+      "loss": 3.5112,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.93544196236413e-05,
+      "loss": 3.5034,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.934603367613078e-05,
+      "loss": 3.5128,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.933766410742399e-05,
+      "loss": 3.5122,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.932927815991347e-05,
+      "loss": 3.5207,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.932089221240295e-05,
+      "loss": 3.5291,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.931250626489243e-05,
+      "loss": 3.5218,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.930413669618564e-05,
+      "loss": 3.4978,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.929575074867512e-05,
+      "loss": 3.5117,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.92873648011646e-05,
+      "loss": 3.5187,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9278978853654084e-05,
+      "loss": 3.514,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9270609284947293e-05,
+      "loss": 3.508,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9262223337436773e-05,
+      "loss": 3.5108,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9253837389926253e-05,
+      "loss": 3.5066,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9245451442415733e-05,
+      "loss": 3.5124,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9237081873708945e-05,
+      "loss": 3.5128,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9228695926198425e-05,
+      "loss": 3.5203,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9220309978687905e-05,
+      "loss": 3.518,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9211924031177385e-05,
+      "loss": 3.5213,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9203554462470594e-05,
+      "loss": 3.5014,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9195168514960074e-05,
+      "loss": 3.512,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9186782567449554e-05,
+      "loss": 3.515,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.917839661993903e-05,
+      "loss": 3.5185,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9170027051232246e-05,
+      "loss": 3.5199,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9161641103721726e-05,
+      "loss": 3.5061,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9153255156211206e-05,
+      "loss": 3.5123,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9144869208700683e-05,
+      "loss": 3.5083,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.91364996399939e-05,
+      "loss": 3.5171,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.912811369248338e-05,
+      "loss": 3.5087,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.911972774497286e-05,
+      "loss": 3.5056,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9111341797462332e-05,
+      "loss": 3.5199,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9102972228755548e-05,
+      "loss": 3.5114,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9094586281245028e-05,
+      "loss": 3.5011,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9086200333734504e-05,
+      "loss": 3.5054,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9077814386223984e-05,
+      "loss": 3.5042,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9069428438713464e-05,
+      "loss": 3.5198,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.906105887000668e-05,
+      "loss": 3.5063,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9052672922496153e-05,
+      "loss": 3.5111,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9044286974985637e-05,
+      "loss": 3.5157,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9035917406278853e-05,
+      "loss": 3.5276,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9027531458768332e-05,
+      "loss": 3.509,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9019145511257806e-05,
+      "loss": 3.5078,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9010759563747286e-05,
+      "loss": 3.5119,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9002373616236766e-05,
+      "loss": 3.5146,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8994004047529978e-05,
+      "loss": 3.5004,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8985618100019458e-05,
+      "loss": 3.5075,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8977232152508938e-05,
+      "loss": 3.5065,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8968846204998418e-05,
+      "loss": 3.5123,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8960476636291627e-05,
+      "loss": 3.5087,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8952090688781107e-05,
+      "loss": 3.5175,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.894370474127059e-05,
+      "loss": 3.5082,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.893531879376007e-05,
+      "loss": 3.5085,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.892694922505328e-05,
+      "loss": 3.5092,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.891856327754276e-05,
+      "loss": 3.5243,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.891017733003224e-05,
+      "loss": 3.5169,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.890179138252172e-05,
+      "loss": 3.508,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8893421813814932e-05,
+      "loss": 3.5022,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.888503586630441e-05,
+      "loss": 3.5179,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.887664991879389e-05,
+      "loss": 3.5112,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.88682803500871e-05,
+      "loss": 3.5171,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.885989440257658e-05,
+      "loss": 3.5071,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.885150845506606e-05,
+      "loss": 3.5224,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.884312250755554e-05,
+      "loss": 3.5146,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8834736560045024e-05,
+      "loss": 3.5158,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8826366991338233e-05,
+      "loss": 3.5108,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8817981043827713e-05,
+      "loss": 3.5196,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8809595096317193e-05,
+      "loss": 3.5102,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8801209148806673e-05,
+      "loss": 3.4995,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8792839580099885e-05,
+      "loss": 3.5253,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8784453632589365e-05,
+      "loss": 3.5131,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8776067685078845e-05,
+      "loss": 3.5147,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8767681737568325e-05,
+      "loss": 3.5112,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8759312168861534e-05,
+      "loss": 3.5001,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.849790573120117,
+      "eval_runtime": 302.3693,
+      "eval_samples_per_second": 1262.003,
+      "eval_steps_per_second": 39.439,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8750926221351014e-05,
+      "loss": 3.5026,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8742540273840494e-05,
+      "loss": 3.5092,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8734154326329978e-05,
+      "loss": 3.5154,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8725784757623187e-05,
+      "loss": 3.5071,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8717398810112667e-05,
+      "loss": 3.5185,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8709012862602147e-05,
+      "loss": 3.5026,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8700626915091627e-05,
+      "loss": 3.5066,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8692240967581107e-05,
+      "loss": 3.5054,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.868387139887432e-05,
+      "loss": 3.5218,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.86754854513638e-05,
+      "loss": 3.5028,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.866709950385328e-05,
+      "loss": 3.5065,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.865871355634276e-05,
+      "loss": 3.5068,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.86503603664397e-05,
+      "loss": 3.5109,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.864197441892918e-05,
+      "loss": 3.5026,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.863358847141866e-05,
+      "loss": 3.5038,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.862520252390814e-05,
+      "loss": 3.4922,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.861681657639762e-05,
+      "loss": 3.5026,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.86084306288871e-05,
+      "loss": 3.5005,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.860004468137658e-05,
+      "loss": 3.4942,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.859165873386606e-05,
+      "loss": 3.5246,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8583289165159273e-05,
+      "loss": 3.5006,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8574903217648753e-05,
+      "loss": 3.5075,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8566517270138233e-05,
+      "loss": 3.5061,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8558131322627713e-05,
+      "loss": 3.5182,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.854976175392092e-05,
+      "loss": 3.4965,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.85413758064104e-05,
+      "loss": 3.504,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8532989858899885e-05,
+      "loss": 3.5059,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8524603911389365e-05,
+      "loss": 3.5063,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8516234342682574e-05,
+      "loss": 3.4986,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8507848395172054e-05,
+      "loss": 3.4987,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8499462447661534e-05,
+      "loss": 3.5103,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8491076500151014e-05,
+      "loss": 3.5099,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8482706931444226e-05,
+      "loss": 3.5126,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8474320983933706e-05,
+      "loss": 3.5093,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8465935036423186e-05,
+      "loss": 3.5001,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8457549088912666e-05,
+      "loss": 3.5119,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8449179520205875e-05,
+      "loss": 3.4902,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8440793572695355e-05,
+      "loss": 3.5114,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843240762518484e-05,
+      "loss": 3.4936,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.842402167767432e-05,
+      "loss": 3.4998,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8415652108967528e-05,
+      "loss": 3.4979,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8407266161457008e-05,
+      "loss": 3.5147,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8398880213946488e-05,
+      "loss": 3.5065,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8390494266435968e-05,
+      "loss": 3.4976,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.838212469772918e-05,
+      "loss": 3.5065,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.837373875021866e-05,
+      "loss": 3.5024,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.836535280270814e-05,
+      "loss": 3.5082,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.835696685519762e-05,
+      "loss": 3.4817,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.834859728649083e-05,
+      "loss": 3.4944,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.834021133898031e-05,
+      "loss": 3.5084,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8331825391469792e-05,
+      "loss": 3.5105,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8323455822763e-05,
+      "loss": 3.4976,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.831506987525248e-05,
+      "loss": 3.4948,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.830668392774196e-05,
+      "loss": 3.5025,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.829829798023144e-05,
+      "loss": 3.4816,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8289928411524654e-05,
+      "loss": 3.514,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8281542464014134e-05,
+      "loss": 3.4989,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8273156516503614e-05,
+      "loss": 3.5051,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8264770568993094e-05,
+      "loss": 3.5012,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8256401000286303e-05,
+      "loss": 3.4896,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8248015052775783e-05,
+      "loss": 3.508,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8239629105265263e-05,
+      "loss": 3.4947,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8231243157754746e-05,
+      "loss": 3.4927,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8222873589047955e-05,
+      "loss": 3.4907,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8214487641537435e-05,
+      "loss": 3.5142,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8206101694026915e-05,
+      "loss": 3.4936,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8197715746516395e-05,
+      "loss": 3.4966,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8189346177809607e-05,
+      "loss": 3.4881,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8180960230299087e-05,
+      "loss": 3.4999,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8172574282788567e-05,
+      "loss": 3.507,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8164188335278047e-05,
+      "loss": 3.5088,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8155818766571256e-05,
+      "loss": 3.5003,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8147432819060736e-05,
+      "loss": 3.5098,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8139046871550216e-05,
+      "loss": 3.512,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8130660924039696e-05,
+      "loss": 3.5037,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812229135533291e-05,
+      "loss": 3.5067,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.811390540782239e-05,
+      "loss": 3.4995,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.810551946031187e-05,
+      "loss": 3.4905,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.809713351280135e-05,
+      "loss": 3.497,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.808876394409456e-05,
+      "loss": 3.5012,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.808037799658404e-05,
+      "loss": 3.5039,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.807199204907352e-05,
+      "loss": 3.517,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8063606101563e-05,
+      "loss": 3.5068,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.805523653285621e-05,
+      "loss": 3.4851,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.804685058534569e-05,
+      "loss": 3.4973,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.803846463783517e-05,
+      "loss": 3.5044,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.803007869032465e-05,
+      "loss": 3.5016,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8021709121617862e-05,
+      "loss": 3.4951,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8013323174107342e-05,
+      "loss": 3.4964,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8004937226596822e-05,
+      "loss": 3.4965,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7996551279086302e-05,
+      "loss": 3.4961,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.798818171037951e-05,
+      "loss": 3.4963,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7979795762868995e-05,
+      "loss": 3.5077,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7971409815358475e-05,
+      "loss": 3.5038,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7963023867847948e-05,
+      "loss": 3.5065,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7954654299141164e-05,
+      "loss": 3.4896,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7946268351630644e-05,
+      "loss": 3.4996,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7937882404120124e-05,
+      "loss": 3.504,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.79294964566096e-05,
+      "loss": 3.5024,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7921126887902816e-05,
+      "loss": 3.5025,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7912740940392296e-05,
+      "loss": 3.4954,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.790435499288177e-05,
+      "loss": 3.4969,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.789596904537125e-05,
+      "loss": 3.4991,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7887599476664465e-05,
+      "loss": 3.5025,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.787921352915395e-05,
+      "loss": 3.4921,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.787082758164342e-05,
+      "loss": 3.4921,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.78624416341329e-05,
+      "loss": 3.5042,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7854072065426117e-05,
+      "loss": 3.4958,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7845686117915597e-05,
+      "loss": 3.4913,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7837300170405074e-05,
+      "loss": 3.4908,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7828914222894554e-05,
+      "loss": 3.489,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.782054465418777e-05,
+      "loss": 3.5096,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7812158706677243e-05,
+      "loss": 3.4907,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7803772759166723e-05,
+      "loss": 3.5014,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7795386811656203e-05,
+      "loss": 3.5004,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.778701724294942e-05,
+      "loss": 3.517,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7778631295438895e-05,
+      "loss": 3.4976,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7770245347928375e-05,
+      "loss": 3.4934,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7761859400417855e-05,
+      "loss": 3.4987,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.775348983171107e-05,
+      "loss": 3.5005,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7745103884200548e-05,
+      "loss": 3.4869,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7736717936690028e-05,
+      "loss": 3.494,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7728331989179507e-05,
+      "loss": 3.4898,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7719962420472717e-05,
+      "loss": 3.5028,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7711576472962197e-05,
+      "loss": 3.4919,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7703190525451676e-05,
+      "loss": 3.5037,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7694804577941156e-05,
+      "loss": 3.4953,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.768643500923437e-05,
+      "loss": 3.4964,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.767804906172385e-05,
+      "loss": 3.4917,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.766966311421333e-05,
+      "loss": 3.5124,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.766127716670281e-05,
+      "loss": 3.5048,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7652907597996018e-05,
+      "loss": 3.4904,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76445216504855e-05,
+      "loss": 3.4927,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.763613570297498e-05,
+      "loss": 3.501,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.762774975546446e-05,
+      "loss": 3.5009,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.761938018675767e-05,
+      "loss": 3.5072,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.761099423924715e-05,
+      "loss": 3.4957,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.760260829173663e-05,
+      "loss": 3.5068,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.759422234422611e-05,
+      "loss": 3.4981,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7585852775519323e-05,
+      "loss": 3.5047,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7577466828008803e-05,
+      "loss": 3.4997,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7569080880498283e-05,
+      "loss": 3.5048,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7560694932987762e-05,
+      "loss": 3.4973,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.755232536428097e-05,
+      "loss": 3.4911,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7543939416770455e-05,
+      "loss": 3.5092,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7535553469259935e-05,
+      "loss": 3.5034,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7527183900553144e-05,
+      "loss": 3.4981,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7518797953042624e-05,
+      "loss": 3.4971,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7510412005532104e-05,
+      "loss": 3.4886,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.850475549697876,
+      "eval_runtime": 305.5394,
+      "eval_samples_per_second": 1248.909,
+      "eval_steps_per_second": 39.029,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7502026058021584e-05,
+      "loss": 3.4918,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7493640110511064e-05,
+      "loss": 3.4915,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7485254163000547e-05,
+      "loss": 3.5049,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7476868215490027e-05,
+      "loss": 3.4914,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7468498646783236e-05,
+      "loss": 3.509,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7460112699272716e-05,
+      "loss": 3.4885,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7451726751762196e-05,
+      "loss": 3.4927,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7443340804251676e-05,
+      "loss": 3.4914,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743497123554489e-05,
+      "loss": 3.5095,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742658528803437e-05,
+      "loss": 3.4914,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741819934052385e-05,
+      "loss": 3.4918,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.740981339301333e-05,
+      "loss": 3.4918,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7401443824306537e-05,
+      "loss": 3.4993,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7393057876796017e-05,
+      "loss": 3.4884,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.73846719292855e-05,
+      "loss": 3.4937,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.737628598177498e-05,
+      "loss": 3.4828,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.736790003426446e-05,
+      "loss": 3.487,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.735951408675394e-05,
+      "loss": 3.4859,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.735112813924342e-05,
+      "loss": 3.4815,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.73427421917329e-05,
+      "loss": 3.5138,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.733437262302611e-05,
+      "loss": 3.4903,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7325986675515593e-05,
+      "loss": 3.4872,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7317600728005073e-05,
+      "loss": 3.4944,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7309214780494553e-05,
+      "loss": 3.5046,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7300845211787762e-05,
+      "loss": 3.4842,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7292459264277242e-05,
+      "loss": 3.4907,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7284073316766722e-05,
+      "loss": 3.4918,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7275687369256202e-05,
+      "loss": 3.4952,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7267317800549414e-05,
+      "loss": 3.4817,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7258931853038894e-05,
+      "loss": 3.4881,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7250545905528374e-05,
+      "loss": 3.494,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7242176336821583e-05,
+      "loss": 3.4961,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7233790389311063e-05,
+      "loss": 3.5044,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7225404441800543e-05,
+      "loss": 3.4953,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7217018494290027e-05,
+      "loss": 3.4871,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7208648925583236e-05,
+      "loss": 3.4994,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7200262978072716e-05,
+      "loss": 3.4811,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7191877030562196e-05,
+      "loss": 3.4965,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7183491083051676e-05,
+      "loss": 3.4821,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7175121514344888e-05,
+      "loss": 3.4844,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7166735566834368e-05,
+      "loss": 3.4885,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7158349619323848e-05,
+      "loss": 3.5009,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.714996367181332e-05,
+      "loss": 3.4948,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7141594103106537e-05,
+      "loss": 3.485,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7133208155596017e-05,
+      "loss": 3.4935,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7124822208085497e-05,
+      "loss": 3.4892,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7116436260574974e-05,
+      "loss": 3.4942,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.710806669186819e-05,
+      "loss": 3.4734,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709968074435767e-05,
+      "loss": 3.4778,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709129479684715e-05,
+      "loss": 3.4961,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7082908849336626e-05,
+      "loss": 3.4953,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7074539280629842e-05,
+      "loss": 3.4857,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7066153333119322e-05,
+      "loss": 3.4854,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7057767385608795e-05,
+      "loss": 3.4893,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7049381438098275e-05,
+      "loss": 3.473,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.704101186939149e-05,
+      "loss": 3.4975,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703262592188097e-05,
+      "loss": 3.4871,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7024239974370447e-05,
+      "loss": 3.4937,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7015854026859927e-05,
+      "loss": 3.4858,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7007484458153143e-05,
+      "loss": 3.4848,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6999098510642623e-05,
+      "loss": 3.4875,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6990712563132096e-05,
+      "loss": 3.4847,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698232661562158e-05,
+      "loss": 3.48,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6973957046914795e-05,
+      "loss": 3.477,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.696557109940427e-05,
+      "loss": 3.4995,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.695718515189375e-05,
+      "loss": 3.4842,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.694879920438323e-05,
+      "loss": 3.4855,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6940429635676444e-05,
+      "loss": 3.476,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.693204368816592e-05,
+      "loss": 3.4864,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.69236577406554e-05,
+      "loss": 3.497,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.691527179314488e-05,
+      "loss": 3.4948,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6906902224438097e-05,
+      "loss": 3.4891,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689851627692757e-05,
+      "loss": 3.4971,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689013032941705e-05,
+      "loss": 3.4987,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6881744381906533e-05,
+      "loss": 3.4873,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6873374813199742e-05,
+      "loss": 3.4938,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6864988865689222e-05,
+      "loss": 3.4887,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6856602918178702e-05,
+      "loss": 3.4811,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6848216970668182e-05,
+      "loss": 3.4886,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6839847401961395e-05,
+      "loss": 3.4847,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6831461454450875e-05,
+      "loss": 3.4909,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6823075506940355e-05,
+      "loss": 3.4998,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6814689559429835e-05,
+      "loss": 3.4945,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6806319990723044e-05,
+      "loss": 3.4713,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6797934043212524e-05,
+      "loss": 3.4864,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6789548095702004e-05,
+      "loss": 3.4908,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6781162148191487e-05,
+      "loss": 3.4896,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6772792579484696e-05,
+      "loss": 3.4839,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6764406631974176e-05,
+      "loss": 3.4787,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6756020684463656e-05,
+      "loss": 3.488,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6747634736953136e-05,
+      "loss": 3.4801,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.673926516824635e-05,
+      "loss": 3.4836,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6730879220735828e-05,
+      "loss": 3.4944,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6722493273225308e-05,
+      "loss": 3.4886,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6714107325714788e-05,
+      "loss": 3.4951,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6705737757007997e-05,
+      "loss": 3.4822,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6697351809497477e-05,
+      "loss": 3.4846,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6688965861986957e-05,
+      "loss": 3.4896,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.668057991447644e-05,
+      "loss": 3.494,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.667221034576965e-05,
+      "loss": 3.49,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.666382439825913e-05,
+      "loss": 3.4831,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.665543845074861e-05,
+      "loss": 3.483,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.664705250323809e-05,
+      "loss": 3.4863,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6638682934531302e-05,
+      "loss": 3.4921,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6630296987020782e-05,
+      "loss": 3.4779,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6621911039510262e-05,
+      "loss": 3.4776,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6613525091999742e-05,
+      "loss": 3.4956,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.660515552329295e-05,
+      "loss": 3.4855,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.659676957578243e-05,
+      "loss": 3.4739,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.658838362827191e-05,
+      "loss": 3.4822,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6579997680761394e-05,
+      "loss": 3.4783,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6571628112054603e-05,
+      "loss": 3.4933,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6563242164544083e-05,
+      "loss": 3.4836,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6554856217033563e-05,
+      "loss": 3.4822,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6546470269523043e-05,
+      "loss": 3.4856,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6538100700816256e-05,
+      "loss": 3.5032,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6529714753305736e-05,
+      "loss": 3.4872,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6521328805795216e-05,
+      "loss": 3.4774,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6512942858284696e-05,
+      "loss": 3.4878,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6504573289577905e-05,
+      "loss": 3.4851,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6496187342067385e-05,
+      "loss": 3.4776,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6487801394556865e-05,
+      "loss": 3.4792,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6479415447046348e-05,
+      "loss": 3.4797,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6471045878339557e-05,
+      "loss": 3.4896,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6462659930829037e-05,
+      "loss": 3.4766,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6454273983318517e-05,
+      "loss": 3.4967,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6445888035807997e-05,
+      "loss": 3.4781,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643751846710121e-05,
+      "loss": 3.4857,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642913251959069e-05,
+      "loss": 3.4792,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642074657208017e-05,
+      "loss": 3.4985,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641236062456965e-05,
+      "loss": 3.494,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6403991055862858e-05,
+      "loss": 3.4847,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6395605108352338e-05,
+      "loss": 3.4758,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6387219160841818e-05,
+      "loss": 3.4906,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.63788332133313e-05,
+      "loss": 3.485,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.637046364462451e-05,
+      "loss": 3.4953,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.636207769711399e-05,
+      "loss": 3.4814,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.635369174960347e-05,
+      "loss": 3.4957,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.634530580209295e-05,
+      "loss": 3.4863,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6336936233386163e-05,
+      "loss": 3.4962,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6328550285875643e-05,
+      "loss": 3.4836,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6320164338365123e-05,
+      "loss": 3.4887,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6311778390854603e-05,
+      "loss": 3.4861,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6303408822147812e-05,
+      "loss": 3.4766,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6295022874637292e-05,
+      "loss": 3.494,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6286636927126772e-05,
+      "loss": 3.4956,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6278250979616252e-05,
+      "loss": 3.4813,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6269881410909464e-05,
+      "loss": 3.4865,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6261495463398944e-05,
+      "loss": 3.4754,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.850656747817993,
+      "eval_runtime": 314.1324,
+      "eval_samples_per_second": 1214.746,
+      "eval_steps_per_second": 37.962,
+      "step": 1450080
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.0022392838238843e+18,
+  "trial_name": null,
+  "trial_params": null
+}