diff --git "a/checkpoint-763190/trainer_state.json" "b/checkpoint-763190/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-763190/trainer_state.json"
@@ -0,0 +1,9045 @@
+{
+  "best_metric": 3.8420088291168213,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/transformer/3/checkpoints/checkpoint-763190",
+  "epoch": 0.025000278439663435,
+  "eval_steps": 10,
+  "global_step": 763190,
+  "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": 10.9262,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8448,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1978,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9733,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8185,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6949,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5933,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5267,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4739,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3928,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3519,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3119,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2679,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2061,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1775,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.146,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1069,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0731,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0519,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0249,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 4.9946,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.966,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9543,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9242,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9154,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8875,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8765,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8582,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8245,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8264,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7969,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7862,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7754,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7678,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7425,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7337,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7324,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7159,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7065,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6884,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6721,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96562908036948e-05,
+      "loss": 4.6661,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 4.6422,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 4.6601,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.6347,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9622763392456454e-05,
+      "loss": 4.635,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9614377444945934e-05,
+      "loss": 4.6263,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9605991497435414e-05,
+      "loss": 4.6213,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 4.6119,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5893,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.5815,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957248046500079e-05,
+      "loss": 4.573,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956409451749027e-05,
+      "loss": 4.5807,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955570856997975e-05,
+      "loss": 4.5585,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5691,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.544,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5374,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5333,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5289,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.525,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949702331620985e-05,
+      "loss": 4.5084,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 4.5036,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5131,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4943,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4876,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 4.4876,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4839,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.466,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4773,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.465,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4566,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4622,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4488,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388055134984274e-05,
+      "loss": 4.4605,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4615,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4511,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4344,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4245,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4141,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4273,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4268,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4242,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.3983,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.396,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.4027,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.3904,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.3958,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.3909,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3938,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.3836,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3786,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3672,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3803,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3646,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3498,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3557,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3683,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3612,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3611,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3394,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3552,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3306,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3469,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3517,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3334,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3284,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3044,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3163,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3181,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3143,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3215,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3028,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3048,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3109,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3139,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3002,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.2986,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.2944,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.2939,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.2981,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.285,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.2811,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.2789,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2744,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2793,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.279,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2854,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.2605,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2738,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.2752,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2633,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2801,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2603,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2722,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2644,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.2423,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.243,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.2468,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2545,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.2552,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2537,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.2437,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.24,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.2326,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.2285,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.2368,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.2482,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.2388,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.221,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.2233,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.19038724899292,
+      "eval_runtime": 610.1579,
+      "eval_samples_per_second": 625.397,
+      "eval_steps_per_second": 19.544,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.2252,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.2202,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.2162,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.2165,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.2069,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.1919,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.1979,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.2106,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.1975,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.2019,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.2037,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.2043,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.1957,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.1966,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.189,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.175,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.1831,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.188,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.1851,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8583249856030314e-05,
+      "loss": 4.1896,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574863908519794e-05,
+      "loss": 4.1737,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.1826,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.1728,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.182,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.1717,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.1693,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.1735,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.1559,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.1716,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.1541,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.1591,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.1601,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.1658,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.846589572729423e-05,
+      "loss": 4.1537,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.1486,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.1617,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.1508,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.1596,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.1428,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.134,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.1377,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.1325,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.1471,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.838206900979649e-05,
+      "loss": 4.1337,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837368306228597e-05,
+      "loss": 4.1398,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.1356,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.1407,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.1386,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834015565104763e-05,
+      "loss": 4.1158,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833176970353711e-05,
+      "loss": 4.1228,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.1172,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.1337,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.1126,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.1327,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.1079,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.1093,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.1178,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.1109,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.1103,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.1066,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.1004,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.1145,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8222801522311535e-05,
+      "loss": 4.1023,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8214415574801015e-05,
+      "loss": 4.1001,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.1009,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.1048,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.086,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.1062,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8172502216052144e-05,
+      "loss": 4.0921,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.0949,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.1,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.0925,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.81389748048138e-05,
+      "loss": 4.107,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.1115,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.1057,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.811383334108597e-05,
+      "loss": 4.0896,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.810544739357545e-05,
+      "loss": 4.0812,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.0788,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.0898,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.0992,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.0934,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.0767,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.0712,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.0837,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.0712,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.0787,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.0813,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.0864,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.0758,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.0739,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.0661,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.0764,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797133774862205e-05,
+      "loss": 4.0747,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796295180111153e-05,
+      "loss": 4.0553,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.0645,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794619628489423e-05,
+      "loss": 4.0748,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793781033738371e-05,
+      "loss": 4.0816,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792942438987319e-05,
+      "loss": 4.0709,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.0609,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7912668873655876e-05,
+      "loss": 4.0737,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7904282926145356e-05,
+      "loss": 4.0521,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7895896978634836e-05,
+      "loss": 4.0692,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7887511031124316e-05,
+      "loss": 4.0745,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7879141462417525e-05,
+      "loss": 4.0638,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7870755514907005e-05,
+      "loss": 4.0642,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7862369567396485e-05,
+      "loss": 4.0422,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7853983619885965e-05,
+      "loss": 4.0443,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845597672375445e-05,
+      "loss": 4.0523,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837211724864925e-05,
+      "loss": 4.0529,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.0603,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.0472,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.78120702611371e-05,
+      "loss": 4.0449,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.0553,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.0625,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.0489,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.0545,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777015690238823e-05,
+      "loss": 4.0463,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.0466,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.053,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77450154386604e-05,
+      "loss": 4.0388,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.0401,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.0383,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.0374,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.0412,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77030857011078e-05,
+      "loss": 4.045,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.0507,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.768631380608676e-05,
+      "loss": 4.0309,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.767794423737997e-05,
+      "loss": 4.038,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.0499,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.0353,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.0483,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7644416826141616e-05,
+      "loss": 4.0335,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7636030878631096e-05,
+      "loss": 4.05,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762764493112058e-05,
+      "loss": 4.0397,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761925898361006e-05,
+      "loss": 4.023,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761088941490327e-05,
+      "loss": 4.023,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760250346739275e-05,
+      "loss": 4.0282,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759411751988223e-05,
+      "loss": 4.0403,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758573157237171e-05,
+      "loss": 4.0351,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757736200366492e-05,
+      "loss": 4.038,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75689760561544e-05,
+      "loss": 4.0311,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756059010864388e-05,
+      "loss": 4.0242,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755220416113336e-05,
+      "loss": 4.0243,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754383459242657e-05,
+      "loss": 4.0186,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.753544864491605e-05,
+      "loss": 4.0264,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7527062697405536e-05,
+      "loss": 4.042,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7518676749895016e-05,
+      "loss": 4.0257,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7510307181188225e-05,
+      "loss": 4.02,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7501921233677705e-05,
+      "loss": 4.0164,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.021358489990234,
+      "eval_runtime": 573.4841,
+      "eval_samples_per_second": 665.391,
+      "eval_steps_per_second": 20.794,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493551664970914e-05,
+      "loss": 4.0235,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485165717460394e-05,
+      "loss": 4.0212,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476779769949874e-05,
+      "loss": 4.013,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468393822439354e-05,
+      "loss": 4.0163,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460007874928834e-05,
+      "loss": 4.014,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451621927418314e-05,
+      "loss": 3.9954,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443235979907794e-05,
+      "loss": 4.0035,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434850032397274e-05,
+      "loss": 4.0169,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426464084886754e-05,
+      "loss": 3.9997,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741809451617997e-05,
+      "loss": 4.0077,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740970856866945e-05,
+      "loss": 4.0156,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740132262115893e-05,
+      "loss": 4.013,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739295305245214e-05,
+      "loss": 4.0024,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738456710494162e-05,
+      "loss": 4.0089,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73761811574311e-05,
+      "loss": 4.0003,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736779520992058e-05,
+      "loss": 3.9933,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735940926241006e-05,
+      "loss": 3.9974,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735102331489954e-05,
+      "loss": 4.0036,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734263736738902e-05,
+      "loss": 4.0044,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73342514198785e-05,
+      "loss": 4.0066,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732586547236797e-05,
+      "loss": 3.9932,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731747952485746e-05,
+      "loss": 4.0049,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730909357734694e-05,
+      "loss": 3.9955,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300724008640154e-05,
+      "loss": 4.0007,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729233806112963e-05,
+      "loss": 3.995,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728395211361911e-05,
+      "loss": 3.9944,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727556616610859e-05,
+      "loss": 4.0008,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72671965974018e-05,
+      "loss": 3.9825,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258810649891277e-05,
+      "loss": 3.9961,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250424702380757e-05,
+      "loss": 3.9847,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242038754870237e-05,
+      "loss": 3.9875,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233669186163446e-05,
+      "loss": 3.9901,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225283238652926e-05,
+      "loss": 3.9941,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721689729114241e-05,
+      "loss": 3.9878,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720851134363189e-05,
+      "loss": 3.9793,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72001417749251e-05,
+      "loss": 3.9938,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719175582741458e-05,
+      "loss": 3.9839,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718336987990406e-05,
+      "loss": 3.9973,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717498393239354e-05,
+      "loss": 3.9811,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716661436368675e-05,
+      "loss": 3.9697,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715822841617623e-05,
+      "loss": 3.9731,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714984246866571e-05,
+      "loss": 3.9746,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714145652115519e-05,
+      "loss": 3.9902,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71330869524484e-05,
+      "loss": 3.9687,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712470100493788e-05,
+      "loss": 3.984,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116315057427366e-05,
+      "loss": 3.9794,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107929109916846e-05,
+      "loss": 3.9778,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099559541210055e-05,
+      "loss": 3.9846,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091173593699535e-05,
+      "loss": 3.9593,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082787646189015e-05,
+      "loss": 3.9662,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074401698678495e-05,
+      "loss": 3.9602,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7066032129971704e-05,
+      "loss": 3.9814,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057646182461184e-05,
+      "loss": 3.9606,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049260234950664e-05,
+      "loss": 3.9795,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040874287440144e-05,
+      "loss": 3.9573,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032488339929624e-05,
+      "loss": 3.9583,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024102392419104e-05,
+      "loss": 3.9637,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015716444908584e-05,
+      "loss": 3.9584,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70073468762018e-05,
+      "loss": 3.9656,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699896092869128e-05,
+      "loss": 3.9626,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699057498118076e-05,
+      "loss": 3.9481,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698218903367024e-05,
+      "loss": 3.9645,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697381946496345e-05,
+      "loss": 3.9562,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696543351745293e-05,
+      "loss": 3.9607,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695704756994241e-05,
+      "loss": 3.9531,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694866162243189e-05,
+      "loss": 3.9598,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69402920537251e-05,
+      "loss": 3.9405,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693190610621458e-05,
+      "loss": 3.9621,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692352015870406e-05,
+      "loss": 3.9489,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691513421119354e-05,
+      "loss": 3.9571,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690676464248675e-05,
+      "loss": 3.9555,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689837869497623e-05,
+      "loss": 3.9515,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688999274746571e-05,
+      "loss": 3.965,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688160679995519e-05,
+      "loss": 3.9718,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68732372312484e-05,
+      "loss": 3.9671,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686485128373788e-05,
+      "loss": 3.9559,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685646533622736e-05,
+      "loss": 3.9443,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684807938871684e-05,
+      "loss": 3.9369,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683970982001005e-05,
+      "loss": 3.9566,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683132387249953e-05,
+      "loss": 3.9623,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682293792498901e-05,
+      "loss": 3.9598,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681455197747849e-05,
+      "loss": 3.9423,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680618240877171e-05,
+      "loss": 3.9382,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679779646126119e-05,
+      "loss": 3.9468,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678941051375067e-05,
+      "loss": 3.9368,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678102456624015e-05,
+      "loss": 3.9441,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772654997533356e-05,
+      "loss": 3.9503,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764269050022836e-05,
+      "loss": 3.9515,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755883102512316e-05,
+      "loss": 3.9443,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747513533805525e-05,
+      "loss": 3.9466,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739127586295005e-05,
+      "loss": 3.9364,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730741638784485e-05,
+      "loss": 3.9473,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722355691273965e-05,
+      "loss": 3.9421,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713969743763445e-05,
+      "loss": 3.9311,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705583796252925e-05,
+      "loss": 3.9362,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669721422754614e-05,
+      "loss": 3.9461,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668882828003562e-05,
+      "loss": 3.9543,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66804423325251e-05,
+      "loss": 3.9418,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667205638501458e-05,
+      "loss": 3.9397,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663670437504054e-05,
+      "loss": 3.9464,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655284489993534e-05,
+      "loss": 3.9267,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6646898542483013e-05,
+      "loss": 3.9445,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638512594972493e-05,
+      "loss": 3.9484,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663015940506944e-05,
+      "loss": 3.9366,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662177345755892e-05,
+      "loss": 3.9453,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66133875100484e-05,
+      "loss": 3.9215,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660500156253788e-05,
+      "loss": 3.919,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659661561502736e-05,
+      "loss": 3.9327,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658822966751684e-05,
+      "loss": 3.9315,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657984372000632e-05,
+      "loss": 3.9407,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65714577724958e-05,
+      "loss": 3.9262,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656308820378901e-05,
+      "loss": 3.9234,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655470225627849e-05,
+      "loss": 3.9304,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654631630876797e-05,
+      "loss": 3.943,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653793036125745e-05,
+      "loss": 3.9323,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529560792550656e-05,
+      "loss": 3.9396,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521174845040136e-05,
+      "loss": 3.9249,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512788897529616e-05,
+      "loss": 3.925,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504402950019096e-05,
+      "loss": 3.9398,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649604976011605e-05,
+      "loss": 3.9217,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648766381260553e-05,
+      "loss": 3.9232,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479277865095e-05,
+      "loss": 3.9222,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647089191758448e-05,
+      "loss": 3.9193,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646250597007396e-05,
+      "loss": 3.9248,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645412002256344e-05,
+      "loss": 3.9264,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644573407505292e-05,
+      "loss": 3.9381,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64373481275424e-05,
+      "loss": 3.9164,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642897855883561e-05,
+      "loss": 3.9204,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642059261132509e-05,
+      "loss": 3.9392,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641220666381457e-05,
+      "loss": 3.9199,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403837095107786e-05,
+      "loss": 3.9367,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395451147597266e-05,
+      "loss": 3.9232,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387065200086746e-05,
+      "loss": 3.9351,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378679252576226e-05,
+      "loss": 3.9274,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370293305065705e-05,
+      "loss": 3.9129,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361907357555185e-05,
+      "loss": 3.9144,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353521410044665e-05,
+      "loss": 3.9147,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345135462534145e-05,
+      "loss": 3.9261,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336765893827354e-05,
+      "loss": 3.9289,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328379946316834e-05,
+      "loss": 3.9246,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319993998806314e-05,
+      "loss": 3.9204,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311608051295794e-05,
+      "loss": 3.9152,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303238482589003e-05,
+      "loss": 3.916,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629485253507848e-05,
+      "loss": 3.909,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628646658756797e-05,
+      "loss": 3.9198,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627808064005745e-05,
+      "loss": 3.9292,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626971107135066e-05,
+      "loss": 3.9176,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626132512384014e-05,
+      "loss": 3.9138,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625295555513335e-05,
+      "loss": 3.9108,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.946960926055908,
+      "eval_runtime": 561.2521,
+      "eval_samples_per_second": 679.892,
+      "eval_steps_per_second": 21.247,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624456960762283e-05,
+      "loss": 3.9148,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623618366011231e-05,
+      "loss": 3.9194,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622779771260179e-05,
+      "loss": 3.9081,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621941176509127e-05,
+      "loss": 3.9091,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621102581758075e-05,
+      "loss": 3.9116,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620263987007023e-05,
+      "loss": 3.8901,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619425392255971e-05,
+      "loss": 3.8972,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618586797504919e-05,
+      "loss": 3.9123,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6177498406342404e-05,
+      "loss": 3.9026,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6169112458831884e-05,
+      "loss": 3.9083,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160726511321364e-05,
+      "loss": 3.9097,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615234056381084e-05,
+      "loss": 3.9069,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614397099510405e-05,
+      "loss": 3.9001,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613558504759353e-05,
+      "loss": 3.9101,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6127199100083006e-05,
+      "loss": 3.8955,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118813152572486e-05,
+      "loss": 3.8933,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110427205061966e-05,
+      "loss": 3.8944,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6102041257551446e-05,
+      "loss": 3.9004,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093655310040926e-05,
+      "loss": 3.9046,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085269362530406e-05,
+      "loss": 3.9084,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607689979382362e-05,
+      "loss": 3.8949,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60685138463131e-05,
+      "loss": 3.9076,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606012789880258e-05,
+      "loss": 3.8985,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605174195129206e-05,
+      "loss": 3.8976,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604335600378154e-05,
+      "loss": 3.8984,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603497005627102e-05,
+      "loss": 3.8927,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60265841087605e-05,
+      "loss": 3.9022,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601819816124998e-05,
+      "loss": 3.8863,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600982859254319e-05,
+      "loss": 3.9018,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600144264503267e-05,
+      "loss": 3.8868,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599305669752215e-05,
+      "loss": 3.8888,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598467075001163e-05,
+      "loss": 3.895,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5976317560108575e-05,
+      "loss": 3.9025,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5967931612598055e-05,
+      "loss": 3.8881,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5959545665087535e-05,
+      "loss": 3.8911,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5951159717577015e-05,
+      "loss": 3.8944,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5942773770066495e-05,
+      "loss": 3.8894,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5934404201359704e-05,
+      "loss": 3.9076,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5926018253849184e-05,
+      "loss": 3.8868,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5917632306338664e-05,
+      "loss": 3.8762,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5909246358828144e-05,
+      "loss": 3.8819,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590087679012135e-05,
+      "loss": 3.8762,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589249084261083e-05,
+      "loss": 3.8977,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588410489510031e-05,
+      "loss": 3.8819,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58757189475898e-05,
+      "loss": 3.8898,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586733300007928e-05,
+      "loss": 3.8842,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585894705256876e-05,
+      "loss": 3.888,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585056110505824e-05,
+      "loss": 3.8913,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584217515754772e-05,
+      "loss": 3.8705,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583380558884093e-05,
+      "loss": 3.8703,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582541964133041e-05,
+      "loss": 3.8748,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581705007262362e-05,
+      "loss": 3.8894,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58086641251131e-05,
+      "loss": 3.8717,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580027817760258e-05,
+      "loss": 3.8859,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579189223009206e-05,
+      "loss": 3.8689,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783522661385267e-05,
+      "loss": 3.8667,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577513671387475e-05,
+      "loss": 3.8775,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576675076636423e-05,
+      "loss": 3.8652,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575836481885371e-05,
+      "loss": 3.8819,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574997887134319e-05,
+      "loss": 3.8763,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57416093026364e-05,
+      "loss": 3.8596,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.573322335512588e-05,
+      "loss": 3.8747,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.572483740761536e-05,
+      "loss": 3.8676,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.571645146010484e-05,
+      "loss": 3.8756,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5708065512594315e-05,
+      "loss": 3.8635,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569969594388753e-05,
+      "loss": 3.8741,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569130999637701e-05,
+      "loss": 3.8572,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568292404886649e-05,
+      "loss": 3.872,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.567453810135597e-05,
+      "loss": 3.8588,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.566615215384545e-05,
+      "loss": 3.8674,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565776620633493e-05,
+      "loss": 3.8704,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564939663762814e-05,
+      "loss": 3.8704,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564101069011762e-05,
+      "loss": 3.8742,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56326247426071e-05,
+      "loss": 3.8881,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562423879509658e-05,
+      "loss": 3.8813,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561586922638979e-05,
+      "loss": 3.8695,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560748327887927e-05,
+      "loss": 3.8628,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559909733136875e-05,
+      "loss": 3.856,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559071138385823e-05,
+      "loss": 3.87,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5582341815151445e-05,
+      "loss": 3.8789,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5573955867640925e-05,
+      "loss": 3.8758,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5565569920130405e-05,
+      "loss": 3.8578,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557183972619885e-05,
+      "loss": 3.8556,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548798025109365e-05,
+      "loss": 3.8688,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540428456402574e-05,
+      "loss": 3.8484,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5532042508892054e-05,
+      "loss": 3.8637,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5523656561381534e-05,
+      "loss": 3.8648,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5515270613871014e-05,
+      "loss": 3.868,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5506884666360494e-05,
+      "loss": 3.8651,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54985150976537e-05,
+      "loss": 3.8682,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549012915014318e-05,
+      "loss": 3.8534,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548174320263266e-05,
+      "loss": 3.8638,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547335725512214e-05,
+      "loss": 3.8587,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546498768641536e-05,
+      "loss": 3.8545,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545660173890484e-05,
+      "loss": 3.8546,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544821579139432e-05,
+      "loss": 3.8694,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54398298438838e-05,
+      "loss": 3.8694,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543144389637328e-05,
+      "loss": 3.8677,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542305794886276e-05,
+      "loss": 3.8574,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541467200135224e-05,
+      "loss": 3.8662,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.540628605384172e-05,
+      "loss": 3.8499,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539791648513493e-05,
+      "loss": 3.8619,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389546916428136e-05,
+      "loss": 3.8686,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5381160968917616e-05,
+      "loss": 3.8562,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372775021407096e-05,
+      "loss": 3.869,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5364389073896576e-05,
+      "loss": 3.839,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535600312638606e-05,
+      "loss": 3.8419,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534763355767927e-05,
+      "loss": 3.854,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533924761016875e-05,
+      "loss": 3.8545,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533086166265823e-05,
+      "loss": 3.8618,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532247571514771e-05,
+      "loss": 3.8477,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531408976763719e-05,
+      "loss": 3.8448,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53057201989304e-05,
+      "loss": 3.853,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529733425141988e-05,
+      "loss": 3.8649,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528894830390936e-05,
+      "loss": 3.8569,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528056235639884e-05,
+      "loss": 3.8625,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527217640888832e-05,
+      "loss": 3.8478,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526380684018153e-05,
+      "loss": 3.8469,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255420892671017e-05,
+      "loss": 3.8651,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5247034945160496e-05,
+      "loss": 3.8467,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238648997649976e-05,
+      "loss": 3.8435,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5230279428943186e-05,
+      "loss": 3.8484,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5221893481432665e-05,
+      "loss": 3.8407,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5213507533922145e-05,
+      "loss": 3.8528,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5205121586411625e-05,
+      "loss": 3.8502,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51967356389011e-05,
+      "loss": 3.8643,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518834969139058e-05,
+      "loss": 3.8417,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5179980122683794e-05,
+      "loss": 3.8425,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5171594175173274e-05,
+      "loss": 3.8678,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5163208227662754e-05,
+      "loss": 3.8427,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5154822280152234e-05,
+      "loss": 3.8577,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5146436332641714e-05,
+      "loss": 3.8522,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5138050385131194e-05,
+      "loss": 3.8599,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51296808164244e-05,
+      "loss": 3.8515,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512129486891388e-05,
+      "loss": 3.8449,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.511290892140336e-05,
+      "loss": 3.838,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510452297389284e-05,
+      "loss": 3.8403,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509615340518605e-05,
+      "loss": 3.8493,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508776745767553e-05,
+      "loss": 3.8571,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507938151016501e-05,
+      "loss": 3.8492,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507099556265449e-05,
+      "loss": 3.8521,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506262599394771e-05,
+      "loss": 3.8406,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505424004643719e-05,
+      "loss": 3.8453,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504585409892667e-05,
+      "loss": 3.8336,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503746815141615e-05,
+      "loss": 3.8493,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502908220390563e-05,
+      "loss": 3.8526,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502071263519884e-05,
+      "loss": 3.8532,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501232668768832e-05,
+      "loss": 3.8393,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50039407401778e-05,
+      "loss": 3.8387,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.905977487564087,
+      "eval_runtime": 573.4819,
+      "eval_samples_per_second": 665.393,
+      "eval_steps_per_second": 20.794,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499555479266728e-05,
+      "loss": 3.847,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498716884515676e-05,
+      "loss": 3.8433,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497878289764624e-05,
+      "loss": 3.8395,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497039695013572e-05,
+      "loss": 3.8392,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962011002625197e-05,
+      "loss": 3.839,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953625055114677e-05,
+      "loss": 3.8243,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945239107604156e-05,
+      "loss": 3.8286,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936853160093636e-05,
+      "loss": 3.8343,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492848359138685e-05,
+      "loss": 3.8348,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492009764387633e-05,
+      "loss": 3.8394,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491171169636581e-05,
+      "loss": 3.8375,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903325748855285e-05,
+      "loss": 3.8331,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48949561801485e-05,
+      "loss": 3.8319,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488657023263798e-05,
+      "loss": 3.8403,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487818428512746e-05,
+      "loss": 3.8297,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869798337616934e-05,
+      "loss": 3.8231,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861412390106414e-05,
+      "loss": 3.8245,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853026442595894e-05,
+      "loss": 3.8353,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844640495085374e-05,
+      "loss": 3.8323,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836254547574854e-05,
+      "loss": 3.8431,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482788497886807e-05,
+      "loss": 3.8271,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481949903135755e-05,
+      "loss": 3.8377,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481111308384703e-05,
+      "loss": 3.8315,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480272713633651e-05,
+      "loss": 3.8304,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479435756762972e-05,
+      "loss": 3.8289,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47859716201192e-05,
+      "loss": 3.8251,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477758567260868e-05,
+      "loss": 3.8342,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476919972509816e-05,
+      "loss": 3.8195,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476083015639137e-05,
+      "loss": 3.8328,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475244420888085e-05,
+      "loss": 3.8227,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474405826137033e-05,
+      "loss": 3.8161,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473567231385981e-05,
+      "loss": 3.83,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727302745153024e-05,
+      "loss": 3.8356,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718916797642504e-05,
+      "loss": 3.8257,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710530850131984e-05,
+      "loss": 3.8222,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702144902621464e-05,
+      "loss": 3.8279,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469377533391467e-05,
+      "loss": 3.8257,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468538938640415e-05,
+      "loss": 3.8402,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467700343889363e-05,
+      "loss": 3.8216,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466861749138311e-05,
+      "loss": 3.8096,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466024792267632e-05,
+      "loss": 3.8167,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46518619751658e-05,
+      "loss": 3.8132,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464347602765528e-05,
+      "loss": 3.8292,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463509008014476e-05,
+      "loss": 3.8173,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462670413263425e-05,
+      "loss": 3.8277,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461833456392746e-05,
+      "loss": 3.8221,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460994861641694e-05,
+      "loss": 3.8203,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460156266890642e-05,
+      "loss": 3.8297,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45931767213959e-05,
+      "loss": 3.8084,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584807152689106e-05,
+      "loss": 3.8052,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576421205178586e-05,
+      "loss": 3.8157,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4568035257668066e-05,
+      "loss": 3.8218,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559649310157546e-05,
+      "loss": 3.8068,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551279741450755e-05,
+      "loss": 3.8217,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542893793940235e-05,
+      "loss": 3.8063,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534507846429715e-05,
+      "loss": 3.8027,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45261218989192e-05,
+      "loss": 3.814,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451775233021241e-05,
+      "loss": 3.802,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450936638270189e-05,
+      "loss": 3.8208,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450098043519137e-05,
+      "loss": 3.8151,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449259448768085e-05,
+      "loss": 3.7943,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448422491897406e-05,
+      "loss": 3.8113,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447583897146354e-05,
+      "loss": 3.8067,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446745302395302e-05,
+      "loss": 3.8185,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44590670764425e-05,
+      "loss": 3.8004,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445069750773571e-05,
+      "loss": 3.8119,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444231156022519e-05,
+      "loss": 3.7962,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443392561271467e-05,
+      "loss": 3.809,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425539665204156e-05,
+      "loss": 3.7996,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4417153717693636e-05,
+      "loss": 3.8029,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408767770183116e-05,
+      "loss": 3.8117,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400398201476325e-05,
+      "loss": 3.8048,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4392012253965805e-05,
+      "loss": 3.8145,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383626306455285e-05,
+      "loss": 3.8245,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4375240358944764e-05,
+      "loss": 3.8225,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4366870790237974e-05,
+      "loss": 3.8094,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4358484842727454e-05,
+      "loss": 3.8024,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4350098895216933e-05,
+      "loss": 3.7957,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341712947706413e-05,
+      "loss": 3.8076,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433334337899962e-05,
+      "loss": 3.8232,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43249574314891e-05,
+      "loss": 3.8117,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431657148397859e-05,
+      "loss": 3.7988,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430818553646807e-05,
+      "loss": 3.7917,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429981596776128e-05,
+      "loss": 3.8133,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429143002025076e-05,
+      "loss": 3.7904,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428304407274024e-05,
+      "loss": 3.8,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427465812522971e-05,
+      "loss": 3.8016,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426628855652293e-05,
+      "loss": 3.8067,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425790260901241e-05,
+      "loss": 3.8081,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424951666150189e-05,
+      "loss": 3.8105,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424113071399136e-05,
+      "loss": 3.7965,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232761145284576e-05,
+      "loss": 3.8042,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224375197774056e-05,
+      "loss": 3.802,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421598925026354e-05,
+      "loss": 3.7951,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207603302753016e-05,
+      "loss": 3.795,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419925011284996e-05,
+      "loss": 3.8122,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419086416533944e-05,
+      "loss": 3.8103,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418247821782892e-05,
+      "loss": 3.811,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41740922703184e-05,
+      "loss": 3.7999,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416570632280788e-05,
+      "loss": 3.805,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415732037529736e-05,
+      "loss": 3.7914,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4148934427786834e-05,
+      "loss": 3.8078,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140548480276314e-05,
+      "loss": 3.8061,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132162532765794e-05,
+      "loss": 3.7993,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412379296405901e-05,
+      "loss": 3.8131,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411540701654849e-05,
+      "loss": 3.7846,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410702106903797e-05,
+      "loss": 3.7861,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409863512152745e-05,
+      "loss": 3.7937,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409026555282066e-05,
+      "loss": 3.7955,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408187960531014e-05,
+      "loss": 3.804,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407349365779962e-05,
+      "loss": 3.7963,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40651077102891e-05,
+      "loss": 3.7828,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405673814158231e-05,
+      "loss": 3.7958,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404835219407179e-05,
+      "loss": 3.808,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403996624656127e-05,
+      "loss": 3.8007,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403158029905075e-05,
+      "loss": 3.8075,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4023210730343963e-05,
+      "loss": 3.7932,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014824782833443e-05,
+      "loss": 3.7899,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4006438835322923e-05,
+      "loss": 3.8083,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39980528878124e-05,
+      "loss": 3.7923,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398968331910561e-05,
+      "loss": 3.7889,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398129737159509e-05,
+      "loss": 3.7892,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397291142408457e-05,
+      "loss": 3.7861,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396452547657405e-05,
+      "loss": 3.7978,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395615590786726e-05,
+      "loss": 3.7955,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394776996035674e-05,
+      "loss": 3.8052,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393938401284622e-05,
+      "loss": 3.7848,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39309980653357e-05,
+      "loss": 3.7887,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392262849662892e-05,
+      "loss": 3.8073,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39142425491184e-05,
+      "loss": 3.791,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390585660160788e-05,
+      "loss": 3.8004,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389747065409736e-05,
+      "loss": 3.8009,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889101085390566e-05,
+      "loss": 3.8011,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880715137880046e-05,
+      "loss": 3.7997,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872329190369526e-05,
+      "loss": 3.786,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863943242859006e-05,
+      "loss": 3.7851,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855573674152215e-05,
+      "loss": 3.7895,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847187726641695e-05,
+      "loss": 3.7937,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838801779131175e-05,
+      "loss": 3.7989,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830415831620655e-05,
+      "loss": 3.7958,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382204626291387e-05,
+      "loss": 3.798,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381366031540335e-05,
+      "loss": 3.7895,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380527436789283e-05,
+      "loss": 3.7906,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379688842038231e-05,
+      "loss": 3.7778,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378851885167552e-05,
+      "loss": 3.7967,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780132904165e-05,
+      "loss": 3.7984,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377174695665448e-05,
+      "loss": 3.7979,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376336100914396e-05,
+      "loss": 3.7826,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375499144043717e-05,
+      "loss": 3.7867,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.881382703781128,
+      "eval_runtime": 571.5891,
+      "eval_samples_per_second": 667.597,
+      "eval_steps_per_second": 20.863,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374662187173038e-05,
+      "loss": 3.8034,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3738235924219865e-05,
+      "loss": 3.7859,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3729849976709345e-05,
+      "loss": 3.7868,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721464029198824e-05,
+      "loss": 3.7839,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3713078081688304e-05,
+      "loss": 3.7856,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704692134177784e-05,
+      "loss": 3.7733,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3696306186667264e-05,
+      "loss": 3.7772,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3687920239156744e-05,
+      "loss": 3.7763,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3679550670449953e-05,
+      "loss": 3.7843,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367118110174316e-05,
+      "loss": 3.7897,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366279515423264e-05,
+      "loss": 3.7821,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365440920672212e-05,
+      "loss": 3.7816,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36460232592116e-05,
+      "loss": 3.7853,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363763731170108e-05,
+      "loss": 3.787,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362925136419056e-05,
+      "loss": 3.7774,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362086541668005e-05,
+      "loss": 3.771,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361247946916953e-05,
+      "loss": 3.773,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360410990046274e-05,
+      "loss": 3.7812,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359572395295222e-05,
+      "loss": 3.7807,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35873380054417e-05,
+      "loss": 3.7918,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357895205793118e-05,
+      "loss": 3.7765,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357058248922439e-05,
+      "loss": 3.7856,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356219654171387e-05,
+      "loss": 3.783,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355381059420335e-05,
+      "loss": 3.7788,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354542464669283e-05,
+      "loss": 3.7771,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3537055077986036e-05,
+      "loss": 3.7762,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528669130475516e-05,
+      "loss": 3.7844,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3520283182964996e-05,
+      "loss": 3.7666,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351189723545448e-05,
+      "loss": 3.7852,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350352766674769e-05,
+      "loss": 3.771,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349514171923717e-05,
+      "loss": 3.7642,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348675577172665e-05,
+      "loss": 3.7807,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347836982421613e-05,
+      "loss": 3.7819,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347000025550934e-05,
+      "loss": 3.7805,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346161430799882e-05,
+      "loss": 3.7737,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34532283604883e-05,
+      "loss": 3.7733,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344484241297778e-05,
+      "loss": 3.7733,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343647284427099e-05,
+      "loss": 3.7905,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342808689676047e-05,
+      "loss": 3.7733,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341970094924995e-05,
+      "loss": 3.7604,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411315001739436e-05,
+      "loss": 3.7673,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402945433032645e-05,
+      "loss": 3.7621,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394559485522125e-05,
+      "loss": 3.7779,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3386173538011605e-05,
+      "loss": 3.7728,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377787590501085e-05,
+      "loss": 3.7736,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369418021794294e-05,
+      "loss": 3.7765,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3361032074283774e-05,
+      "loss": 3.7722,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352646126773254e-05,
+      "loss": 3.7818,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3344260179262734e-05,
+      "loss": 3.7601,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333589061055594e-05,
+      "loss": 3.7541,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332750466304542e-05,
+      "loss": 3.7702,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33191187155349e-05,
+      "loss": 3.7656,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331073276802439e-05,
+      "loss": 3.762,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33023631993176e-05,
+      "loss": 3.7734,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329397725180708e-05,
+      "loss": 3.7563,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328559130429656e-05,
+      "loss": 3.757,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327720535678604e-05,
+      "loss": 3.7649,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326883578807925e-05,
+      "loss": 3.755,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326044984056873e-05,
+      "loss": 3.7709,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325206389305821e-05,
+      "loss": 3.7634,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324367794554768e-05,
+      "loss": 3.7553,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32353083768409e-05,
+      "loss": 3.7596,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322692242933038e-05,
+      "loss": 3.7542,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321853648181986e-05,
+      "loss": 3.7739,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321015053430934e-05,
+      "loss": 3.7559,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320178096560255e-05,
+      "loss": 3.7674,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319339501809203e-05,
+      "loss": 3.7481,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3185009070581506e-05,
+      "loss": 3.759,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3176623123070986e-05,
+      "loss": 3.7527,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31682535543642e-05,
+      "loss": 3.7543,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315986760685368e-05,
+      "loss": 3.7613,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151481659343155e-05,
+      "loss": 3.7593,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3143095711832635e-05,
+      "loss": 3.7654,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313472614312585e-05,
+      "loss": 3.7764,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312634019561533e-05,
+      "loss": 3.775,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311795424810481e-05,
+      "loss": 3.7658,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310956830059429e-05,
+      "loss": 3.757,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101198731887506e-05,
+      "loss": 3.7521,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309281278437698e-05,
+      "loss": 3.7585,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308442683686646e-05,
+      "loss": 3.7756,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307604088935594e-05,
+      "loss": 3.765,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067671320649155e-05,
+      "loss": 3.7542,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305928537313863e-05,
+      "loss": 3.7435,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305089942562811e-05,
+      "loss": 3.7705,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304251347811759e-05,
+      "loss": 3.7387,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034143909410804e-05,
+      "loss": 3.7571,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025757961900284e-05,
+      "loss": 3.7569,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3017372014389764e-05,
+      "loss": 3.761,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008986066879244e-05,
+      "loss": 3.7606,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300061649817245e-05,
+      "loss": 3.7668,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299223055066193e-05,
+      "loss": 3.7472,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298384460315141e-05,
+      "loss": 3.7564,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297545865564089e-05,
+      "loss": 3.7587,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29670890869341e-05,
+      "loss": 3.7495,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295870313942358e-05,
+      "loss": 3.7502,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29503335707168e-05,
+      "loss": 3.766,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294194762320628e-05,
+      "loss": 3.7622,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293356167569576e-05,
+      "loss": 3.7706,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292517572818524e-05,
+      "loss": 3.7538,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291678978067472e-05,
+      "loss": 3.7575,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29084038331642e-05,
+      "loss": 3.7447,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290001788565368e-05,
+      "loss": 3.7605,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289163193814316e-05,
+      "loss": 3.7633,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288326236943637e-05,
+      "loss": 3.7513,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287487642192585e-05,
+      "loss": 3.7657,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286649047441533e-05,
+      "loss": 3.7388,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285810452690481e-05,
+      "loss": 3.7447,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849734958198016e-05,
+      "loss": 3.7443,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2841349010687496e-05,
+      "loss": 3.7521,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2832963063176976e-05,
+      "loss": 3.7602,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824577115666456e-05,
+      "loss": 3.7554,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281620754695967e-05,
+      "loss": 3.7306,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280782159944915e-05,
+      "loss": 3.7568,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279943565193863e-05,
+      "loss": 3.7611,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279104970442811e-05,
+      "loss": 3.7597,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278268013572132e-05,
+      "loss": 3.7582,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27742941882108e-05,
+      "loss": 3.7497,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276590824070028e-05,
+      "loss": 3.7457,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275752229318976e-05,
+      "loss": 3.7659,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274915272448297e-05,
+      "loss": 3.7481,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274076677697245e-05,
+      "loss": 3.7476,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273238082946193e-05,
+      "loss": 3.7426,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272399488195141e-05,
+      "loss": 3.7462,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715625313244625e-05,
+      "loss": 3.7547,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2707239365734105e-05,
+      "loss": 3.7526,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698853418223585e-05,
+      "loss": 3.761,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690467470713065e-05,
+      "loss": 3.7459,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2682097902006274e-05,
+      "loss": 3.7435,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673711954495754e-05,
+      "loss": 3.7613,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2665326006985234e-05,
+      "loss": 3.7457,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656940059474714e-05,
+      "loss": 3.7557,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264857049076792e-05,
+      "loss": 3.7595,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26401845432574e-05,
+      "loss": 3.7588,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263179859574688e-05,
+      "loss": 3.7558,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262341264823636e-05,
+      "loss": 3.7446,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261504307952958e-05,
+      "loss": 3.7372,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260665713201906e-05,
+      "loss": 3.7473,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259827118450854e-05,
+      "loss": 3.7511,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258988523699802e-05,
+      "loss": 3.7601,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258151566829123e-05,
+      "loss": 3.7517,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257312972078071e-05,
+      "loss": 3.7515,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256474377327019e-05,
+      "loss": 3.75,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255635782575967e-05,
+      "loss": 3.751,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254798825705288e-05,
+      "loss": 3.7387,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253960230954236e-05,
+      "loss": 3.7456,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253121636203184e-05,
+      "loss": 3.7587,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252283041452132e-05,
+      "loss": 3.7568,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251446084581453e-05,
+      "loss": 3.741,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250607489830401e-05,
+      "loss": 3.7413,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8656201362609863,
+      "eval_runtime": 575.9172,
+      "eval_samples_per_second": 662.58,
+      "eval_steps_per_second": 20.706,
+      "step": 457914
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249770532959722e-05,
+      "loss": 3.7627,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24893193820867e-05,
+      "loss": 3.7451,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248093343457618e-05,
+      "loss": 3.7414,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247254748706566e-05,
+      "loss": 3.7452,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246416153955514e-05,
+      "loss": 3.7387,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245577559204462e-05,
+      "loss": 3.7331,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24473896445341e-05,
+      "loss": 3.7332,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243900369702358e-05,
+      "loss": 3.7367,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243063412831679e-05,
+      "loss": 3.744,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242224818080627e-05,
+      "loss": 3.7451,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413878612099486e-05,
+      "loss": 3.742,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405492664588966e-05,
+      "loss": 3.7348,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2397106717078446e-05,
+      "loss": 3.7491,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388720769567926e-05,
+      "loss": 3.7428,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380351200861135e-05,
+      "loss": 3.7408,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371965253350615e-05,
+      "loss": 3.7267,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363579305840095e-05,
+      "loss": 3.7297,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2355193358329575e-05,
+      "loss": 3.744,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346807410819055e-05,
+      "loss": 3.7381,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233842146330853e-05,
+      "loss": 3.7505,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233003551579801e-05,
+      "loss": 3.7341,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232164956828749e-05,
+      "loss": 3.7466,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2313279999580704e-05,
+      "loss": 3.741,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230491043087392e-05,
+      "loss": 3.7371,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22965244833634e-05,
+      "loss": 3.7358,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228813853585288e-05,
+      "loss": 3.7349,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227975258834236e-05,
+      "loss": 3.7427,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227136664083183e-05,
+      "loss": 3.7295,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226298069332131e-05,
+      "loss": 3.7434,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225459474581079e-05,
+      "loss": 3.7341,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224620879830027e-05,
+      "loss": 3.7217,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223783922959348e-05,
+      "loss": 3.7395,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222945328208296e-05,
+      "loss": 3.7409,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222106733457244e-05,
+      "loss": 3.7443,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221268138706193e-05,
+      "loss": 3.7306,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220431181835514e-05,
+      "loss": 3.7363,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219592587084462e-05,
+      "loss": 3.7323,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21875399233341e-05,
+      "loss": 3.7464,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217915397582358e-05,
+      "loss": 3.7361,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170784407116787e-05,
+      "loss": 3.7209,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2162398459606267e-05,
+      "loss": 3.7262,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2154012512095746e-05,
+      "loss": 3.7204,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145626564585226e-05,
+      "loss": 3.7414,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2137256995878436e-05,
+      "loss": 3.7319,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128871048367915e-05,
+      "loss": 3.7329,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120485100857395e-05,
+      "loss": 3.7392,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211209915334688e-05,
+      "loss": 3.7291,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210372958464009e-05,
+      "loss": 3.7416,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209534363712957e-05,
+      "loss": 3.7214,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208695768961905e-05,
+      "loss": 3.7159,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207858812091226e-05,
+      "loss": 3.729,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207020217340174e-05,
+      "loss": 3.7234,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206181622589122e-05,
+      "loss": 3.7273,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20534302783807e-05,
+      "loss": 3.7303,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204506070967391e-05,
+      "loss": 3.7229,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203667476216339e-05,
+      "loss": 3.718,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202828881465287e-05,
+      "loss": 3.7232,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201990286714235e-05,
+      "loss": 3.7201,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011533298435565e-05,
+      "loss": 3.7302,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2003147350925045e-05,
+      "loss": 3.7247,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994761403414525e-05,
+      "loss": 3.7147,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1986391834707734e-05,
+      "loss": 3.722,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1978005887197214e-05,
+      "loss": 3.7098,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969619939686694e-05,
+      "loss": 3.7386,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1961233992176174e-05,
+      "loss": 3.7192,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1952848044665654e-05,
+      "loss": 3.7254,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1944462097155134e-05,
+      "loss": 3.7144,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1936076149644614e-05,
+      "loss": 3.7136,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927690202134094e-05,
+      "loss": 3.7175,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19193206334273e-05,
+      "loss": 3.7162,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191093468591679e-05,
+      "loss": 3.7251,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190254873840627e-05,
+      "loss": 3.7203,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189416279089575e-05,
+      "loss": 3.7282,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188579322218896e-05,
+      "loss": 3.7372,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187740727467844e-05,
+      "loss": 3.7332,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186902132716792e-05,
+      "loss": 3.7309,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18606353796574e-05,
+      "loss": 3.7191,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185226581095061e-05,
+      "loss": 3.7137,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184387986344009e-05,
+      "loss": 3.7209,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835510294733297e-05,
+      "loss": 3.7353,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1827124347222776e-05,
+      "loss": 3.727,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818738399712256e-05,
+      "loss": 3.7168,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181035245220174e-05,
+      "loss": 3.7037,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180196650469122e-05,
+      "loss": 3.7347,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17935805571807e-05,
+      "loss": 3.6986,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178519460967018e-05,
+      "loss": 3.7181,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177682504096339e-05,
+      "loss": 3.7224,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176843909345287e-05,
+      "loss": 3.721,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176005314594235e-05,
+      "loss": 3.7277,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175166719843183e-05,
+      "loss": 3.7311,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174329762972504e-05,
+      "loss": 3.7067,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173491168221452e-05,
+      "loss": 3.7214,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726525734704e-05,
+      "loss": 3.7223,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171813978719348e-05,
+      "loss": 3.7136,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170977021848669e-05,
+      "loss": 3.7107,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701400649779906e-05,
+      "loss": 3.7301,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1693014702269386e-05,
+      "loss": 3.7258,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684628754758866e-05,
+      "loss": 3.7361,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1676242807248346e-05,
+      "loss": 3.7139,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667856859737826e-05,
+      "loss": 3.7231,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1659470912227306e-05,
+      "loss": 3.7084,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1651084964716786e-05,
+      "loss": 3.7231,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164269901720626e-05,
+      "loss": 3.7265,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1634329448499475e-05,
+      "loss": 3.7221,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625943500988955e-05,
+      "loss": 3.7266,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617557553478435e-05,
+      "loss": 3.7045,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609171605967915e-05,
+      "loss": 3.7043,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160081841606486e-05,
+      "loss": 3.7048,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159243246855434e-05,
+      "loss": 3.7151,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158404652104382e-05,
+      "loss": 3.7233,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15756605735333e-05,
+      "loss": 3.7228,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156727462602278e-05,
+      "loss": 3.6953,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155888867851226e-05,
+      "loss": 3.7186,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155050273100173e-05,
+      "loss": 3.7249,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154211678349121e-05,
+      "loss": 3.7234,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153374721478443e-05,
+      "loss": 3.7213,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152536126727391e-05,
+      "loss": 3.7136,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151697531976338e-05,
+      "loss": 3.7129,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150858937225287e-05,
+      "loss": 3.7248,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1500219803546084e-05,
+      "loss": 3.7154,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149183385603556e-05,
+      "loss": 3.711,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148344790852504e-05,
+      "loss": 3.705,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147506196101452e-05,
+      "loss": 3.708,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146669239230773e-05,
+      "loss": 3.7181,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1458306444797206e-05,
+      "loss": 3.7164,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449920497286686e-05,
+      "loss": 3.7226,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1441534549776166e-05,
+      "loss": 3.7111,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143316498106938e-05,
+      "loss": 3.707,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424779033558855e-05,
+      "loss": 3.7268,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1416393086048335e-05,
+      "loss": 3.7067,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140800713853782e-05,
+      "loss": 3.7212,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139963756983103e-05,
+      "loss": 3.7271,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139125162232051e-05,
+      "loss": 3.7193,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138286567480999e-05,
+      "loss": 3.7203,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137449610610321e-05,
+      "loss": 3.7142,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136611015859268e-05,
+      "loss": 3.7005,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135772421108216e-05,
+      "loss": 3.7094,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134933826357164e-05,
+      "loss": 3.7126,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134095231606112e-05,
+      "loss": 3.7254,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13325663685506e-05,
+      "loss": 3.7126,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132418042104008e-05,
+      "loss": 3.7205,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131581085233329e-05,
+      "loss": 3.712,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307424904822776e-05,
+      "loss": 3.7162,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1299038957312256e-05,
+      "loss": 3.7015,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290653009801736e-05,
+      "loss": 3.7106,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1282283441094945e-05,
+      "loss": 3.7248,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273897493584425e-05,
+      "loss": 3.722,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1265511546073905e-05,
+      "loss": 3.7041,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1257125598563384e-05,
+      "loss": 3.7107,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.855581283569336,
+      "eval_runtime": 571.6992,
+      "eval_samples_per_second": 667.468,
+      "eval_steps_per_second": 20.859,
+      "step": 534233
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1248756029856594e-05,
+      "loss": 3.7299,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1240370082346074e-05,
+      "loss": 3.7082,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1231984134835553e-05,
+      "loss": 3.7068,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1223598187325033e-05,
+      "loss": 3.7139,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1215212239814513e-05,
+      "loss": 3.6996,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120682629230399e-05,
+      "loss": 3.7022,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119844034479347e-05,
+      "loss": 3.695,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119005439728296e-05,
+      "loss": 3.7054,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118168482857617e-05,
+      "loss": 3.7056,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117329888106565e-05,
+      "loss": 3.7119,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116492931235886e-05,
+      "loss": 3.7072,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115654336484834e-05,
+      "loss": 3.7021,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114815741733782e-05,
+      "loss": 3.7143,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11397714698273e-05,
+      "loss": 3.7049,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113138552231678e-05,
+      "loss": 3.7114,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112299957480626e-05,
+      "loss": 3.6888,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111461362729574e-05,
+      "loss": 3.6973,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110624405858895e-05,
+      "loss": 3.7072,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109785811107843e-05,
+      "loss": 3.7016,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108947216356791e-05,
+      "loss": 3.7179,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1081086216057394e-05,
+      "loss": 3.703,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10727166473506e-05,
+      "loss": 3.7086,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106433069984008e-05,
+      "loss": 3.707,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105596113113329e-05,
+      "loss": 3.6998,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104757518362277e-05,
+      "loss": 3.7058,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103918923611225e-05,
+      "loss": 3.6975,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103080328860173e-05,
+      "loss": 3.7112,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102241734109121e-05,
+      "loss": 3.6992,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1014031393580685e-05,
+      "loss": 3.7066,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1005645446070165e-05,
+      "loss": 3.6982,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0997259498559645e-05,
+      "loss": 3.6905,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098888992985286e-05,
+      "loss": 3.7056,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098050398234234e-05,
+      "loss": 3.7067,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097211803483182e-05,
+      "loss": 3.7073,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09637320873213e-05,
+      "loss": 3.7026,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0955378897418245e-05,
+      "loss": 3.6981,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946992949907725e-05,
+      "loss": 3.7002,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0938607002397205e-05,
+      "loss": 3.7119,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0930221054886685e-05,
+      "loss": 3.705,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092183510737616e-05,
+      "loss": 3.6911,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913465538669374e-05,
+      "loss": 3.691,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0905079591158854e-05,
+      "loss": 3.6878,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896693643648334e-05,
+      "loss": 3.7067,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0888307696137814e-05,
+      "loss": 3.7005,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879921748627294e-05,
+      "loss": 3.6946,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0871535801116774e-05,
+      "loss": 3.7096,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0863149853606254e-05,
+      "loss": 3.6965,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085478028489946e-05,
+      "loss": 3.709,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084639433738894e-05,
+      "loss": 3.6883,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083800838987842e-05,
+      "loss": 3.6839,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08296224423679e-05,
+      "loss": 3.6935,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082123649485738e-05,
+      "loss": 3.6947,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081285054734686e-05,
+      "loss": 3.6897,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080446459983634e-05,
+      "loss": 3.6985,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079609503112955e-05,
+      "loss": 3.6922,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078770908361904e-05,
+      "loss": 3.6843,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077932313610852e-05,
+      "loss": 3.6953,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0770937188598e-05,
+      "loss": 3.6816,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076255124108748e-05,
+      "loss": 3.6976,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075416529357696e-05,
+      "loss": 3.6923,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074579572487017e-05,
+      "loss": 3.6854,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073740977735965e-05,
+      "loss": 3.6889,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072902382984913e-05,
+      "loss": 3.6797,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072063788233861e-05,
+      "loss": 3.7038,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071225193482809e-05,
+      "loss": 3.685,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07038823661213e-05,
+      "loss": 3.6932,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0695496418610777e-05,
+      "loss": 3.6853,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0687110471100257e-05,
+      "loss": 3.6819,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0678724523589736e-05,
+      "loss": 3.6842,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067033857607922e-05,
+      "loss": 3.6862,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06619526285687e-05,
+      "loss": 3.69,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0653566681058176e-05,
+      "loss": 3.6902,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0645180733547656e-05,
+      "loss": 3.6969,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063681116484087e-05,
+      "loss": 3.7039,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0628425217330345e-05,
+      "loss": 3.7026,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0620039269819825e-05,
+      "loss": 3.6939,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0611653322309305e-05,
+      "loss": 3.6928,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060328375360252e-05,
+      "loss": 3.6766,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594897806091994e-05,
+      "loss": 3.6892,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058652823738521e-05,
+      "loss": 3.7039,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057814228987469e-05,
+      "loss": 3.6939,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056975634236418e-05,
+      "loss": 3.6911,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056137039485365e-05,
+      "loss": 3.6723,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055298444734313e-05,
+      "loss": 3.6968,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0544614878636346e-05,
+      "loss": 3.6713,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053622893112582e-05,
+      "loss": 3.6852,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05278429836153e-05,
+      "loss": 3.686,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051945703610478e-05,
+      "loss": 3.6894,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051107108859426e-05,
+      "loss": 3.6969,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050268514108374e-05,
+      "loss": 3.6948,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049429919357322e-05,
+      "loss": 3.6804,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04859132460627e-05,
+      "loss": 3.6917,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0477543677355915e-05,
+      "loss": 3.686,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0469157729845395e-05,
+      "loss": 3.683,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460771782334875e-05,
+      "loss": 3.6798,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0452385834824355e-05,
+      "loss": 3.6942,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0444016266117564e-05,
+      "loss": 3.6947,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0435630318607044e-05,
+      "loss": 3.7025,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0427244371096524e-05,
+      "loss": 3.6895,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418858423586004e-05,
+      "loss": 3.6901,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0410472476075483e-05,
+      "loss": 3.6771,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0402086528564963e-05,
+      "loss": 3.695,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039370058105444e-05,
+      "loss": 3.6927,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038533101234765e-05,
+      "loss": 3.6912,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037694506483713e-05,
+      "loss": 3.6936,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036855911732661e-05,
+      "loss": 3.6764,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036017316981609e-05,
+      "loss": 3.6736,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035180360110931e-05,
+      "loss": 3.6766,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034341765359879e-05,
+      "loss": 3.6826,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033503170608827e-05,
+      "loss": 3.6927,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032664575857775e-05,
+      "loss": 3.6918,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031825981106723e-05,
+      "loss": 3.6615,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030989024236044e-05,
+      "loss": 3.691,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030150429484992e-05,
+      "loss": 3.6942,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02931183473394e-05,
+      "loss": 3.6914,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028473239982888e-05,
+      "loss": 3.69,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027634645231836e-05,
+      "loss": 3.6878,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026796050480783e-05,
+      "loss": 3.6764,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025957455729732e-05,
+      "loss": 3.6955,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025120498859053e-05,
+      "loss": 3.6855,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024281904108001e-05,
+      "loss": 3.6784,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0234433093569486e-05,
+      "loss": 3.6718,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0226047146058966e-05,
+      "loss": 3.6816,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217661198548446e-05,
+      "loss": 3.6864,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0209275251037926e-05,
+      "loss": 3.6825,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0200889303527406e-05,
+      "loss": 3.6933,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0192519734820615e-05,
+      "loss": 3.6812,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018415016611383e-05,
+      "loss": 3.6776,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175764218603304e-05,
+      "loss": 3.6972,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0167378271092784e-05,
+      "loss": 3.6737,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015899232358227e-05,
+      "loss": 3.6905,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015060637607175e-05,
+      "loss": 3.6965,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014222042856123e-05,
+      "loss": 3.6853,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013383448105071e-05,
+      "loss": 3.6914,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012546491234392e-05,
+      "loss": 3.6833,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01170789648334e-05,
+      "loss": 3.6711,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010869301732288e-05,
+      "loss": 3.6777,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010030706981236e-05,
+      "loss": 3.6853,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009192112230184e-05,
+      "loss": 3.6918,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008353517479132e-05,
+      "loss": 3.6859,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007516560608453e-05,
+      "loss": 3.6903,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006677965857401e-05,
+      "loss": 3.6797,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005839371106349e-05,
+      "loss": 3.6867,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005000776355297e-05,
+      "loss": 3.6718,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0041621816042455e-05,
+      "loss": 3.6812,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0033235868531935e-05,
+      "loss": 3.693,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0024849921021415e-05,
+      "loss": 3.6941,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0016463973510895e-05,
+      "loss": 3.6723,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0008094404804104e-05,
+      "loss": 3.6837,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8487296104431152,
+      "eval_runtime": 589.3545,
+      "eval_samples_per_second": 647.473,
+      "eval_steps_per_second": 20.234,
+      "step": 610552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999972483609731e-05,
+      "loss": 3.7002,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999133888858679e-05,
+      "loss": 3.682,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998295294107627e-05,
+      "loss": 3.6762,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997456699356575e-05,
+      "loss": 3.6839,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996618104605523e-05,
+      "loss": 3.6691,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995779509854471e-05,
+      "loss": 3.6716,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994940915103419e-05,
+      "loss": 3.6717,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994102320352367e-05,
+      "loss": 3.6711,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993265363481689e-05,
+      "loss": 3.6742,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99242840661101e-05,
+      "loss": 3.6829,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991589811859958e-05,
+      "loss": 3.6762,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990751217108906e-05,
+      "loss": 3.6748,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989912622357854e-05,
+      "loss": 3.6858,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989074027606802e-05,
+      "loss": 3.6709,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882370707361227e-05,
+      "loss": 3.6852,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873984759850707e-05,
+      "loss": 3.6605,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865598812340187e-05,
+      "loss": 3.6675,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857212864829666e-05,
+      "loss": 3.6787,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848826917319146e-05,
+      "loss": 3.6727,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9840440969808626e-05,
+      "loss": 3.6884,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9832055022298106e-05,
+      "loss": 3.6756,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823669074787586e-05,
+      "loss": 3.6763,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9815299506080795e-05,
+      "loss": 3.6804,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980692993737401e-05,
+      "loss": 3.6703,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979854398986349e-05,
+      "loss": 3.6773,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790158042352964e-05,
+      "loss": 3.6684,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781772094842444e-05,
+      "loss": 3.6809,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773386147331924e-05,
+      "loss": 3.6698,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9765000199821404e-05,
+      "loss": 3.6798,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9756614252310884e-05,
+      "loss": 3.6678,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9748228304800364e-05,
+      "loss": 3.6599,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973985873609358e-05,
+      "loss": 3.6778,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973147278858306e-05,
+      "loss": 3.6788,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972308684107254e-05,
+      "loss": 3.678,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971471727236575e-05,
+      "loss": 3.6717,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970633132485523e-05,
+      "loss": 3.6746,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969794537734471e-05,
+      "loss": 3.669,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968955942983419e-05,
+      "loss": 3.6816,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968117348232367e-05,
+      "loss": 3.6764,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967278753481315e-05,
+      "loss": 3.6616,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966440158730263e-05,
+      "loss": 3.6638,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965601563979211e-05,
+      "loss": 3.6597,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964766244988905e-05,
+      "loss": 3.6777,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9639276502378534e-05,
+      "loss": 3.676,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9630890554868014e-05,
+      "loss": 3.6625,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9622504607357494e-05,
+      "loss": 3.6826,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9614118659846974e-05,
+      "loss": 3.666,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960574909114018e-05,
+      "loss": 3.6794,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959736314362966e-05,
+      "loss": 3.6647,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958897719611914e-05,
+      "loss": 3.6577,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958059124860862e-05,
+      "loss": 3.6637,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95722053010981e-05,
+      "loss": 3.6622,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956381935358758e-05,
+      "loss": 3.6692,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955543340607706e-05,
+      "loss": 3.6632,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954706383737027e-05,
+      "loss": 3.6658,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953867788985975e-05,
+      "loss": 3.6525,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953029194234923e-05,
+      "loss": 3.6669,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952190599483872e-05,
+      "loss": 3.6558,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95135200473282e-05,
+      "loss": 3.6678,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950513409981768e-05,
+      "loss": 3.6586,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949676453111089e-05,
+      "loss": 3.6631,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948837858360037e-05,
+      "loss": 3.6612,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947999263608985e-05,
+      "loss": 3.6513,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947160668857933e-05,
+      "loss": 3.6729,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94632207410688e-05,
+      "loss": 3.6579,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454851172362016e-05,
+      "loss": 3.6656,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9446465224851496e-05,
+      "loss": 3.6601,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943807927734097e-05,
+      "loss": 3.6517,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429693329830456e-05,
+      "loss": 3.6554,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9421307382319936e-05,
+      "loss": 3.659,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941293781361315e-05,
+      "loss": 3.6627,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9404551866102625e-05,
+      "loss": 3.6617,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9396165918592105e-05,
+      "loss": 3.6658,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387779971081585e-05,
+      "loss": 3.6819,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9379394023571065e-05,
+      "loss": 3.6745,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9371008076060545e-05,
+      "loss": 3.6653,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362638507353754e-05,
+      "loss": 3.665,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9354252559843234e-05,
+      "loss": 3.6517,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9345866612332714e-05,
+      "loss": 3.6594,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9337480664822194e-05,
+      "loss": 3.6729,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329094717311674e-05,
+      "loss": 3.6706,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932072514860489e-05,
+      "loss": 3.6595,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931233920109437e-05,
+      "loss": 3.6482,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930395325358385e-05,
+      "loss": 3.6688,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929556730607333e-05,
+      "loss": 3.6405,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928718135856281e-05,
+      "loss": 3.6606,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927879541105229e-05,
+      "loss": 3.6604,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927040946354177e-05,
+      "loss": 3.6593,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926203989483498e-05,
+      "loss": 3.6715,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925365394732446e-05,
+      "loss": 3.6668,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924526799981394e-05,
+      "loss": 3.6551,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923688205230342e-05,
+      "loss": 3.6588,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92284961047929e-05,
+      "loss": 3.6591,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922011015728238e-05,
+      "loss": 3.6562,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921172420977186e-05,
+      "loss": 3.6535,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920333826226134e-05,
+      "loss": 3.6648,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9194968693554554e-05,
+      "loss": 3.6706,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9186582746044034e-05,
+      "loss": 3.678,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9178196798533514e-05,
+      "loss": 3.6577,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916981085102299e-05,
+      "loss": 3.6639,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916142490351247e-05,
+      "loss": 3.653,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915303895600195e-05,
+      "loss": 3.6622,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914466938729516e-05,
+      "loss": 3.6693,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9136283439784636e-05,
+      "loss": 3.6659,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127897492274116e-05,
+      "loss": 3.6681,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9119511544763596e-05,
+      "loss": 3.6478,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9111125597253076e-05,
+      "loss": 3.6489,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102739649742556e-05,
+      "loss": 3.6502,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909437008103577e-05,
+      "loss": 3.6585,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908598413352525e-05,
+      "loss": 3.6595,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907759818601473e-05,
+      "loss": 3.6648,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906921223850421e-05,
+      "loss": 3.6338,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906084266979742e-05,
+      "loss": 3.6639,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90524567222869e-05,
+      "loss": 3.6689,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904407077477638e-05,
+      "loss": 3.6605,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903568482726586e-05,
+      "loss": 3.664,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902729887975534e-05,
+      "loss": 3.6617,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901891293224482e-05,
+      "loss": 3.648,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90105269847343e-05,
+      "loss": 3.667,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900215741602751e-05,
+      "loss": 3.6636,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8993771468516996e-05,
+      "loss": 3.6495,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8985385521006476e-05,
+      "loss": 3.6476,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8976999573495956e-05,
+      "loss": 3.6547,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8968613625985436e-05,
+      "loss": 3.6569,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8960227678474916e-05,
+      "loss": 3.6554,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8951858109768125e-05,
+      "loss": 3.664,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8943472162257605e-05,
+      "loss": 3.6575,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8935086214747085e-05,
+      "loss": 3.6492,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926700267236565e-05,
+      "loss": 3.672,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8918314319726045e-05,
+      "loss": 3.6479,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890992837221552e-05,
+      "loss": 3.6652,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8901558803508734e-05,
+      "loss": 3.6714,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8893172855998214e-05,
+      "loss": 3.6594,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884786908487694e-05,
+      "loss": 3.6646,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8876400960977174e-05,
+      "loss": 3.6584,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886803139227039e-05,
+      "loss": 3.6438,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885964544475987e-05,
+      "loss": 3.6546,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885125949724935e-05,
+      "loss": 3.6548,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884287354973882e-05,
+      "loss": 3.6671,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88344876022283e-05,
+      "loss": 3.6602,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882610165471778e-05,
+      "loss": 3.6594,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881771570720726e-05,
+      "loss": 3.6586,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880932975969674e-05,
+      "loss": 3.6579,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880096019098995e-05,
+      "loss": 3.6442,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879257424347943e-05,
+      "loss": 3.6553,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878418829596892e-05,
+      "loss": 3.6654,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87758023484584e-05,
+      "loss": 3.6665,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876743277975161e-05,
+      "loss": 3.6479,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875904683224109e-05,
+      "loss": 3.6531,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.844803810119629,
+      "eval_runtime": 573.814,
+      "eval_samples_per_second": 665.008,
+      "eval_steps_per_second": 20.782,
+      "step": 686871
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8750677263534297e-05,
+      "loss": 3.6743,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8742291316023777e-05,
+      "loss": 3.6552,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8733905368513256e-05,
+      "loss": 3.6528,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8725519421002736e-05,
+      "loss": 3.6571,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8717133473492216e-05,
+      "loss": 3.6482,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8708747525981696e-05,
+      "loss": 3.6444,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8700377957274905e-05,
+      "loss": 3.6448,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8691992009764385e-05,
+      "loss": 3.6448,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8683606062253865e-05,
+      "loss": 3.6466,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867522011474335e-05,
+      "loss": 3.6586,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866685054603656e-05,
+      "loss": 3.6481,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865846459852604e-05,
+      "loss": 3.6498,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865007865101552e-05,
+      "loss": 3.6594,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8641692703505e-05,
+      "loss": 3.6468,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.863332313479821e-05,
+      "loss": 3.6553,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.862493718728769e-05,
+      "loss": 3.6344,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861655123977717e-05,
+      "loss": 3.6404,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.860816529226665e-05,
+      "loss": 3.6556,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859977934475613e-05,
+      "loss": 3.649,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859140977604934e-05,
+      "loss": 3.6616,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858302382853882e-05,
+      "loss": 3.6448,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8574637881028306e-05,
+      "loss": 3.6601,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8566251933517786e-05,
+      "loss": 3.6478,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8557882364810995e-05,
+      "loss": 3.645,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8549496417300475e-05,
+      "loss": 3.6521,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8541110469789955e-05,
+      "loss": 3.6472,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8532724522279435e-05,
+      "loss": 3.6543,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8524338574768915e-05,
+      "loss": 3.6451,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8515952627258395e-05,
+      "loss": 3.6539,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8507566679747875e-05,
+      "loss": 3.6454,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8499180732237354e-05,
+      "loss": 3.6276,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8490811163530564e-05,
+      "loss": 3.6552,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8482425216020044e-05,
+      "loss": 3.6543,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8474039268509524e-05,
+      "loss": 3.6523,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8465653320999003e-05,
+      "loss": 3.6487,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845728375229222e-05,
+      "loss": 3.6451,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84488978047817e-05,
+      "loss": 3.6459,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844052823607491e-05,
+      "loss": 3.6539,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843214228856439e-05,
+      "loss": 3.6529,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842375634105387e-05,
+      "loss": 3.6338,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841537039354335e-05,
+      "loss": 3.6424,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840700082483656e-05,
+      "loss": 3.637,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839861487732604e-05,
+      "loss": 3.6469,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839022892981552e-05,
+      "loss": 3.6515,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8381842982305e-05,
+      "loss": 3.6397,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837347341359821e-05,
+      "loss": 3.6552,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836508746608769e-05,
+      "loss": 3.6396,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835670151857717e-05,
+      "loss": 3.6563,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834831557106665e-05,
+      "loss": 3.64,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833992962355613e-05,
+      "loss": 3.6325,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8331543676045606e-05,
+      "loss": 3.6413,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8323157728535086e-05,
+      "loss": 3.6332,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314771781024566e-05,
+      "loss": 3.6476,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8306402212317775e-05,
+      "loss": 3.6369,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829803264361099e-05,
+      "loss": 3.6453,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828964669610047e-05,
+      "loss": 3.6281,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828126074858995e-05,
+      "loss": 3.6417,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827287480107943e-05,
+      "loss": 3.6283,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826448885356891e-05,
+      "loss": 3.6468,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825610290605839e-05,
+      "loss": 3.6348,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82477333373516e-05,
+      "loss": 3.6371,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823934738984108e-05,
+      "loss": 3.6369,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823096144233056e-05,
+      "loss": 3.6266,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.822257549482004e-05,
+      "loss": 3.6456,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821420592611325e-05,
+      "loss": 3.6372,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820581997860273e-05,
+      "loss": 3.6442,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819743403109221e-05,
+      "loss": 3.6311,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818904808358169e-05,
+      "loss": 3.6305,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818066213607117e-05,
+      "loss": 3.6299,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817227618856065e-05,
+      "loss": 3.6375,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8163890241050135e-05,
+      "loss": 3.6337,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8155504293539615e-05,
+      "loss": 3.6393,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8147134724832824e-05,
+      "loss": 3.6412,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8138748777322304e-05,
+      "loss": 3.6585,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8130362829811784e-05,
+      "loss": 3.6463,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8121976882301264e-05,
+      "loss": 3.6435,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811360731359447e-05,
+      "loss": 3.6397,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810522136608395e-05,
+      "loss": 3.6299,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809683541857343e-05,
+      "loss": 3.6305,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808844947106291e-05,
+      "loss": 3.6492,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808007990235612e-05,
+      "loss": 3.6496,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80716939548456e-05,
+      "loss": 3.6399,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806330800733509e-05,
+      "loss": 3.6226,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805492205982457e-05,
+      "loss": 3.6378,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804655249111778e-05,
+      "loss": 3.6199,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803816654360726e-05,
+      "loss": 3.6369,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802978059609674e-05,
+      "loss": 3.6359,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802139464858622e-05,
+      "loss": 3.6342,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80130087010757e-05,
+      "loss": 3.646,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800462275356518e-05,
+      "loss": 3.6411,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799625318485839e-05,
+      "loss": 3.6344,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798786723734787e-05,
+      "loss": 3.6355,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797948128983735e-05,
+      "loss": 3.6369,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797109534232683e-05,
+      "loss": 3.629,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796270939481631e-05,
+      "loss": 3.6315,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795432344730579e-05,
+      "loss": 3.638,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7945953878599e-05,
+      "loss": 3.6466,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793756793108848e-05,
+      "loss": 3.6525,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792918198357796e-05,
+      "loss": 3.6327,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920796036067436e-05,
+      "loss": 3.6417,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7912410088556916e-05,
+      "loss": 3.6315,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7904024141046396e-05,
+      "loss": 3.6395,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789565457233961e-05,
+      "loss": 3.6382,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7887268624829085e-05,
+      "loss": 3.6448,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7878882677318565e-05,
+      "loss": 3.6447,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7870496729808045e-05,
+      "loss": 3.6243,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786212716110126e-05,
+      "loss": 3.6264,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785374121359074e-05,
+      "loss": 3.6254,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784535526608022e-05,
+      "loss": 3.6302,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78369693185697e-05,
+      "loss": 3.638,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782858337105918e-05,
+      "loss": 3.641,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782021380235239e-05,
+      "loss": 3.616,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781182785484187e-05,
+      "loss": 3.6379,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780344190733135e-05,
+      "loss": 3.6429,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779505595982083e-05,
+      "loss": 3.6384,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778667001231031e-05,
+      "loss": 3.64,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777828406479979e-05,
+      "loss": 3.6367,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7769914496093e-05,
+      "loss": 3.6265,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776152854858248e-05,
+      "loss": 3.6412,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775314260107196e-05,
+      "loss": 3.6397,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7744756653561445e-05,
+      "loss": 3.6333,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7736370706050925e-05,
+      "loss": 3.6185,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7728001137344134e-05,
+      "loss": 3.6341,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7719615189833614e-05,
+      "loss": 3.6313,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7711229242323094e-05,
+      "loss": 3.6352,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7702843294812574e-05,
+      "loss": 3.6336,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7694457347302054e-05,
+      "loss": 3.6404,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768608777859526e-05,
+      "loss": 3.6264,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767770183108474e-05,
+      "loss": 3.6469,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766931588357422e-05,
+      "loss": 3.6256,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76609299360637e-05,
+      "loss": 3.6414,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765256036735691e-05,
+      "loss": 3.6487,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76441744198464e-05,
+      "loss": 3.6347,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763578847233588e-05,
+      "loss": 3.6412,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762740252482536e-05,
+      "loss": 3.6372,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761903295611857e-05,
+      "loss": 3.6213,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761064700860805e-05,
+      "loss": 3.6304,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760226106109753e-05,
+      "loss": 3.6274,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759387511358701e-05,
+      "loss": 3.6425,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758548916607649e-05,
+      "loss": 3.638,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757710321856597e-05,
+      "loss": 3.6375,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7568733649859176e-05,
+      "loss": 3.6406,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7560347702348656e-05,
+      "loss": 3.6317,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7551961754838136e-05,
+      "loss": 3.6212,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7543575807327616e-05,
+      "loss": 3.6319,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7535189859817096e-05,
+      "loss": 3.6441,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7526803912306576e-05,
+      "loss": 3.6409,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7518417964796056e-05,
+      "loss": 3.6242,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751004839608927e-05,
+      "loss": 3.6291,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8420088291168213,
+      "eval_runtime": 320.9317,
+      "eval_samples_per_second": 1189.01,
+      "eval_steps_per_second": 37.157,
+      "step": 763190
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 5.306040521162035e+17,
+  "trial_name": null,
+  "trial_params": null
+}