diff --git "a/checkpoint-686880/trainer_state.json" "b/checkpoint-686880/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-686880/trainer_state.json"
@@ -0,0 +1,8143 @@
+{
+  "best_metric": 3.848252058029175,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-reconstruction/transformer/0/checkpoints/checkpoint-686880",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 686880,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 11.0716,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8178,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1789,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9638,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8051,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7015,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5961,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5295,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.443,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.394,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3468,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3168,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2648,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.207,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1786,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1281,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1107,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0866,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0455,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0206,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0064,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9814,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9536,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9364,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9265,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8883,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8745,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8662,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8361,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8155,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8067,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740117521192533e-05,
+      "loss": 4.7988,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731731573682013e-05,
+      "loss": 4.7873,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7693,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7594,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7544,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7355,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 4.7073,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 4.7047,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 4.6781,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6881,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6735,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6737,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 4.651,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.6525,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6378,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6303,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9605991497435414e-05,
+      "loss": 4.6343,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 4.5939,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95892359812181e-05,
+      "loss": 4.6027,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.6044,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5979,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.5743,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5678,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5581,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5511,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5665,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5278,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5439,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5351,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5147,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5243,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5009,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4971,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4965,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5061,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4754,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.477,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4611,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4721,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4763,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4722,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4518,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4632,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4676,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4437,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4387,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4312,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4246,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4315,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4118,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4329,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4246,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.409,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3917,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.4022,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.4009,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.4046,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3887,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.3909,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3855,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3725,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3701,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3764,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 4.3696,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3706,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.356,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3548,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3616,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3614,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3526,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3477,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3397,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.337,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3388,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3149,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3259,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3287,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3291,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3154,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.323,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3009,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3287,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.2987,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3094,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3084,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.3127,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.3036,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2941,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.296,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2977,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.286,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.2855,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2803,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.3006,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2688,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2736,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2704,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2746,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2656,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2825,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2711,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2652,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2564,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2692,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.2632,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.268,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.251,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2638,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.257,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2593,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2392,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2586,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2413,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2246,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2548,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2397,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2392,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2391,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.2253,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.196438789367676,
+      "eval_runtime": 307.708,
+      "eval_samples_per_second": 1240.107,
+      "eval_steps_per_second": 38.754,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.2071,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.214,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2336,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2178,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2242,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2149,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.2083,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.2023,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2068,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.2046,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.2181,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.2123,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.1969,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.1897,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.1837,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.1822,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.1937,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.1828,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1803,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.1983,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1854,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.1805,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.1826,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1941,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.1654,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.1697,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.1795,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.1687,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1574,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1576,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.1713,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.1664,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1643,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1609,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845757529499864e-05,
+      "loss": 4.1624,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844918934748812e-05,
+      "loss": 4.1604,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1454,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1496,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1348,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1481,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.1425,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.1545,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1399,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1465,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.1411,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.1406,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1511,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1178,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1294,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.144,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1388,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1253,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83066937550242e-05,
+      "loss": 4.1196,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829832418631741e-05,
+      "loss": 4.1202,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828993823880689e-05,
+      "loss": 4.1098,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828155229129637e-05,
+      "loss": 4.1345,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.1082,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1276,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.1209,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1081,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.1152,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231252985036985e-05,
+      "loss": 4.1036,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222867037526465e-05,
+      "loss": 4.1026,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.106,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820609514250542e-05,
+      "loss": 4.1236,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.094,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189339626288114e-05,
+      "loss": 4.1017,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.0848,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.1011,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.115,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1075,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.0981,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813904032002872e-05,
+      "loss": 4.108,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81306543725182e-05,
+      "loss": 4.1127,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.0948,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.0925,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810551290879037e-05,
+      "loss": 4.0932,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809712696127985e-05,
+      "loss": 4.0881,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808874101376933e-05,
+      "loss": 4.0934,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.0855,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807198549755203e-05,
+      "loss": 4.0997,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806359955004151e-05,
+      "loss": 4.0941,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805521360253099e-05,
+      "loss": 4.0892,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804682765502047e-05,
+      "loss": 4.071,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803844170750995e-05,
+      "loss": 4.0817,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803005575999943e-05,
+      "loss": 4.0886,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021686191292636e-05,
+      "loss": 4.0901,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013300243782116e-05,
+      "loss": 4.0832,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004914296271596e-05,
+      "loss": 4.0795,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996528348761076e-05,
+      "loss": 4.084,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988158780054285e-05,
+      "loss": 4.0721,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979772832543765e-05,
+      "loss": 4.0664,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971386885033245e-05,
+      "loss": 4.0825,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963000937522725e-05,
+      "loss": 4.0743,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795461499001221e-05,
+      "loss": 4.0787,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794624542130542e-05,
+      "loss": 4.0725,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79378594737949e-05,
+      "loss": 4.0628,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792947352628438e-05,
+      "loss": 4.0769,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792108757877386e-05,
+      "loss": 4.0814,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791271801006707e-05,
+      "loss": 4.0692,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.0665,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789594611504603e-05,
+      "loss": 4.0588,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788756016753551e-05,
+      "loss": 4.0642,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787919059882872e-05,
+      "loss": 4.0714,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708046513182e-05,
+      "loss": 4.0482,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786241870380768e-05,
+      "loss": 4.0549,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854032756297166e-05,
+      "loss": 4.0606,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845663187590375e-05,
+      "loss": 4.0682,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837293618883584e-05,
+      "loss": 4.052,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828907671373064e-05,
+      "loss": 4.0605,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820521723862544e-05,
+      "loss": 4.0471,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812135776352024e-05,
+      "loss": 4.0681,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803749828841504e-05,
+      "loss": 4.0478,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795363881330984e-05,
+      "loss": 4.0544,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786977933820464e-05,
+      "loss": 4.0627,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778591986309944e-05,
+      "loss": 4.0598,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777023879640689e-05,
+      "loss": 4.0565,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776185284889637e-05,
+      "loss": 4.0447,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775346690138585e-05,
+      "loss": 4.0508,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774508095387533e-05,
+      "loss": 4.0552,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773669500636481e-05,
+      "loss": 4.0425,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772830905885429e-05,
+      "loss": 4.0452,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771992311134377e-05,
+      "loss": 4.0411,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771153716383325e-05,
+      "loss": 4.06,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770316759512646e-05,
+      "loss": 4.0337,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769478164761594e-05,
+      "loss": 4.0418,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768639570010542e-05,
+      "loss": 4.0369,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76780097525949e-05,
+      "loss": 4.0436,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7669640183888106e-05,
+      "loss": 4.0344,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7661254236377586e-05,
+      "loss": 4.0538,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765286828886707e-05,
+      "loss": 4.0493,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764448234135655e-05,
+      "loss": 4.0373,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763611277264976e-05,
+      "loss": 4.0296,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762772682513924e-05,
+      "loss": 4.0455,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761934087762872e-05,
+      "loss": 4.0428,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610954930118195e-05,
+      "loss": 4.051,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760258536141141e-05,
+      "loss": 4.0259,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759419941390089e-05,
+      "loss": 4.0466,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75858298451941e-05,
+      "loss": 4.038,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757744389768358e-05,
+      "loss": 4.049,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756905795017306e-05,
+      "loss": 4.0288,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756067200266254e-05,
+      "loss": 4.04,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755228605515202e-05,
+      "loss": 4.0316,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75439001076415e-05,
+      "loss": 4.0206,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753551416013098e-05,
+      "loss": 4.0414,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752712821262046e-05,
+      "loss": 4.0337,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751875864391367e-05,
+      "loss": 4.0305,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751037269640315e-05,
+      "loss": 4.0272,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750198674889263e-05,
+      "loss": 4.0228,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0265703201293945,
+      "eval_runtime": 314.4532,
+      "eval_samples_per_second": 1213.507,
+      "eval_steps_per_second": 37.923,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749360080138211e-05,
+      "loss": 4.0037,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748521485387159e-05,
+      "loss": 4.0115,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747682890636107e-05,
+      "loss": 4.0349,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746844295885055e-05,
+      "loss": 4.0202,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746005701134003e-05,
+      "loss": 4.0281,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745167106382951e-05,
+      "loss": 4.0129,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744328511631899e-05,
+      "loss": 4.0132,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434899168808475e-05,
+      "loss": 4.0075,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426513221297955e-05,
+      "loss": 4.0132,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418127273787435e-05,
+      "loss": 4.0168,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740974132627691e-05,
+      "loss": 4.0213,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401371757570124e-05,
+      "loss": 4.0233,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739301856766706e-05,
+      "loss": 4.0082,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738463262015654e-05,
+      "loss": 4.0021,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737624667264602e-05,
+      "loss": 4.0025,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73678607251355e-05,
+      "loss": 3.992,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735947477762498e-05,
+      "loss": 4.0056,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735108883011446e-05,
+      "loss": 3.9989,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734270288260394e-05,
+      "loss": 3.9985,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733431693509343e-05,
+      "loss": 4.0203,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732593098758291e-05,
+      "loss": 4.0024,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731754504007238e-05,
+      "loss": 4.0041,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730915909256186e-05,
+      "loss": 4.0073,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730078952385508e-05,
+      "loss": 4.0162,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729240357634456e-05,
+      "loss": 3.9866,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728401762883403e-05,
+      "loss": 3.9979,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727563168132351e-05,
+      "loss": 4.0037,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726726211261673e-05,
+      "loss": 3.9973,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725887616510621e-05,
+      "loss": 3.9875,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725049021759568e-05,
+      "loss": 3.9823,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 4.0031,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 4.0017,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225348753867856e-05,
+      "loss": 3.9925,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216962806357336e-05,
+      "loss": 3.9971,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208576858846816e-05,
+      "loss": 3.9938,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200190911336296e-05,
+      "loss": 3.9915,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191804963825775e-05,
+      "loss": 3.9811,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183419016315255e-05,
+      "loss": 3.9909,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175049447608465e-05,
+      "loss": 3.9686,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166663500097945e-05,
+      "loss": 3.9866,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158277552587424e-05,
+      "loss": 3.983,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149891605076904e-05,
+      "loss": 3.9946,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714152203637012e-05,
+      "loss": 3.9815,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71331360888596e-05,
+      "loss": 3.9834,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712475014134908e-05,
+      "loss": 3.9876,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711636419383856e-05,
+      "loss": 3.9818,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710799462513177e-05,
+      "loss": 3.993,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709960867762125e-05,
+      "loss": 3.9611,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709123910891446e-05,
+      "loss": 3.9783,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708285316140394e-05,
+      "loss": 3.9846,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707446721389342e-05,
+      "loss": 3.9865,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70660812663829e-05,
+      "loss": 3.9713,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705771169767611e-05,
+      "loss": 3.9643,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704932575016559e-05,
+      "loss": 3.9697,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040939802655074e-05,
+      "loss": 3.9562,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032553855144554e-05,
+      "loss": 3.9865,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024167907634034e-05,
+      "loss": 3.964,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015781960123514e-05,
+      "loss": 3.9738,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007396012612994e-05,
+      "loss": 3.9735,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6999010065102474e-05,
+      "loss": 3.9616,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699064049639568e-05,
+      "loss": 3.9679,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698225454888516e-05,
+      "loss": 3.9591,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697386860137464e-05,
+      "loss": 3.9582,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696548265386412e-05,
+      "loss": 3.957,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695711308515733e-05,
+      "loss": 3.9771,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694872713764681e-05,
+      "loss": 3.9563,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694035756894003e-05,
+      "loss": 3.956,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693197162142951e-05,
+      "loss": 3.9433,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692358567391899e-05,
+      "loss": 3.9573,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.968,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690681377889795e-05,
+      "loss": 3.9712,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689842783138743e-05,
+      "loss": 3.9582,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689004188387691e-05,
+      "loss": 3.9684,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688165593636639e-05,
+      "loss": 3.9749,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873286367659596e-05,
+      "loss": 3.9588,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864900420149076e-05,
+      "loss": 3.9495,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856530851442285e-05,
+      "loss": 3.9594,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848144903931765e-05,
+      "loss": 3.9525,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839758956421245e-05,
+      "loss": 3.9554,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831373008910725e-05,
+      "loss": 3.9485,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822987061400205e-05,
+      "loss": 3.9599,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681460111388969e-05,
+      "loss": 3.9655,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806215166379165e-05,
+      "loss": 3.9523,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679784559767238e-05,
+      "loss": 3.9397,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678947602896559e-05,
+      "loss": 3.9464,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678109008145507e-05,
+      "loss": 3.9518,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677270413394455e-05,
+      "loss": 3.9582,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676431818643403e-05,
+      "loss": 3.9514,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675593223892351e-05,
+      "loss": 3.9537,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674754629141298e-05,
+      "loss": 3.9519,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673916034390246e-05,
+      "loss": 3.9413,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673077439639194e-05,
+      "loss": 3.9402,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672240482768516e-05,
+      "loss": 3.9458,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714035258978375e-05,
+      "loss": 3.9483,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705649311467855e-05,
+      "loss": 3.9523,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697263363957335e-05,
+      "loss": 3.9467,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688877416446815e-05,
+      "loss": 3.9331,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668049146893629e-05,
+      "loss": 3.95,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667210552142577e-05,
+      "loss": 3.9564,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666371957391525e-05,
+      "loss": 3.9431,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665535000520846e-05,
+      "loss": 3.9471,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664696405769794e-05,
+      "loss": 3.933,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663857811018742e-05,
+      "loss": 3.9404,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66301921626769e-05,
+      "loss": 3.9488,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621806215166383e-05,
+      "loss": 3.924,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613420267655863e-05,
+      "loss": 3.9328,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660505069894907e-05,
+      "loss": 3.9418,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659666475143855e-05,
+      "loss": 3.9405,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658827880392803e-05,
+      "loss": 3.9295,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657989285641751e-05,
+      "loss": 3.9406,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657150690890699e-05,
+      "loss": 3.9261,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656312096139647e-05,
+      "loss": 3.9449,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655473501388595e-05,
+      "loss": 3.9307,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654636544517916e-05,
+      "loss": 3.9314,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653797949766864e-05,
+      "loss": 3.9445,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652959355015812e-05,
+      "loss": 3.9429,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652122398145134e-05,
+      "loss": 3.9388,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651283803394082e-05,
+      "loss": 3.9288,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65044520864303e-05,
+      "loss": 3.9328,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649606613891978e-05,
+      "loss": 3.9346,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648768019140926e-05,
+      "loss": 3.929,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647929424389874e-05,
+      "loss": 3.9321,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647090829638822e-05,
+      "loss": 3.9236,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462538727681426e-05,
+      "loss": 3.9382,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454152780170906e-05,
+      "loss": 3.9279,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445766832660386e-05,
+      "loss": 3.9266,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437380885149866e-05,
+      "loss": 3.9194,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429011316443075e-05,
+      "loss": 3.9315,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420625368932555e-05,
+      "loss": 3.9176,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412239421422035e-05,
+      "loss": 3.9406,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640385347391152e-05,
+      "loss": 3.9364,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395467526401e-05,
+      "loss": 3.9288,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387081578890475e-05,
+      "loss": 3.9134,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378695631379955e-05,
+      "loss": 3.9363,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370309683869435e-05,
+      "loss": 3.9312,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636194011516265e-05,
+      "loss": 3.9351,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635357054645586e-05,
+      "loss": 3.92,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634518459894534e-05,
+      "loss": 3.9294,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633679865143482e-05,
+      "loss": 3.9322,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632841270392429e-05,
+      "loss": 3.9399,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632004313521751e-05,
+      "loss": 3.9218,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631165718770699e-05,
+      "loss": 3.9273,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303271240196475e-05,
+      "loss": 3.9243,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629488529268595e-05,
+      "loss": 3.9114,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628649934517543e-05,
+      "loss": 3.9331,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627811339766491e-05,
+      "loss": 3.9306,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626974382895812e-05,
+      "loss": 3.9188,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62613578814476e-05,
+      "loss": 3.9214,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625297193393708e-05,
+      "loss": 3.9157,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.951601982116699,
+      "eval_runtime": 310.1749,
+      "eval_samples_per_second": 1230.245,
+      "eval_steps_per_second": 38.446,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624458598642656e-05,
+      "loss": 3.9098,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623620003891604e-05,
+      "loss": 3.9035,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227830470209246e-05,
+      "loss": 3.9284,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219444522698726e-05,
+      "loss": 3.9124,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621105857518821e-05,
+      "loss": 3.9278,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620267262767769e-05,
+      "loss": 3.9082,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619428668016717e-05,
+      "loss": 3.9129,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618591711146038e-05,
+      "loss": 3.9006,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617753116394986e-05,
+      "loss": 3.9131,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616914521643934e-05,
+      "loss": 3.9109,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616075926892882e-05,
+      "loss": 3.9191,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615238970022203e-05,
+      "loss": 3.9175,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614402013151524e-05,
+      "loss": 3.9071,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613563418400472e-05,
+      "loss": 3.901,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61272482364942e-05,
+      "loss": 3.9005,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611886228898368e-05,
+      "loss": 3.8934,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611047634147316e-05,
+      "loss": 3.9024,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610209039396265e-05,
+      "loss": 3.9,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609370444645213e-05,
+      "loss": 3.8959,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085318498941607e-05,
+      "loss": 3.9252,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076948930234816e-05,
+      "loss": 3.904,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068562982724296e-05,
+      "loss": 3.9047,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060177035213776e-05,
+      "loss": 3.9106,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051791087703256e-05,
+      "loss": 3.9134,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043405140192735e-05,
+      "loss": 3.8915,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035035571485945e-05,
+      "loss": 3.8981,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026649623975425e-05,
+      "loss": 3.9079,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018263676464904e-05,
+      "loss": 3.9029,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009877728954384e-05,
+      "loss": 3.8874,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001491781443864e-05,
+      "loss": 3.8861,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599312221273708e-05,
+      "loss": 3.9057,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598473626522656e-05,
+      "loss": 3.9075,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597635031771604e-05,
+      "loss": 3.9003,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596796437020552e-05,
+      "loss": 3.9,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595959480149873e-05,
+      "loss": 3.8955,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595120885398821e-05,
+      "loss": 3.8983,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594282290647769e-05,
+      "loss": 3.8839,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593443695896717e-05,
+      "loss": 3.8997,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592606739026038e-05,
+      "loss": 3.8775,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591768144274986e-05,
+      "loss": 3.8911,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590929549523934e-05,
+      "loss": 3.8853,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590090954772882e-05,
+      "loss": 3.9046,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58925236002183e-05,
+      "loss": 3.8933,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884154031511514e-05,
+      "loss": 3.8889,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875768084000994e-05,
+      "loss": 3.8933,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867382136490474e-05,
+      "loss": 3.8948,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858996188979954e-05,
+      "loss": 3.8978,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585061024146943e-05,
+      "loss": 3.871,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584225705156637e-05,
+      "loss": 3.8849,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583387110405585e-05,
+      "loss": 3.8968,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582548515654533e-05,
+      "loss": 3.8937,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581709920903481e-05,
+      "loss": 3.8822,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580871326152429e-05,
+      "loss": 3.8786,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580032731401377e-05,
+      "loss": 3.8794,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579194136650325e-05,
+      "loss": 3.8683,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578355541899273e-05,
+      "loss": 3.8934,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577516947148221e-05,
+      "loss": 3.8732,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576679990277543e-05,
+      "loss": 3.8868,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57584139552649e-05,
+      "loss": 3.8863,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575002800775438e-05,
+      "loss": 3.8711,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574164206024386e-05,
+      "loss": 3.8774,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733272491537076e-05,
+      "loss": 3.8721,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572488654402655e-05,
+      "loss": 3.869,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571650059651603e-05,
+      "loss": 3.8758,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570811464900551e-05,
+      "loss": 3.8853,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699745080298725e-05,
+      "loss": 3.87,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691359132788205e-05,
+      "loss": 3.8698,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682973185277685e-05,
+      "loss": 3.8592,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674587237767165e-05,
+      "loss": 3.8699,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666217669060374e-05,
+      "loss": 3.8787,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657831721549854e-05,
+      "loss": 3.8907,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649445774039334e-05,
+      "loss": 3.8698,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641059826528814e-05,
+      "loss": 3.8831,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632673879018294e-05,
+      "loss": 3.8861,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56243043103115e-05,
+      "loss": 3.8822,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561591836280098e-05,
+      "loss": 3.8606,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560753241529046e-05,
+      "loss": 3.88,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559914646777994e-05,
+      "loss": 3.8726,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559077689907316e-05,
+      "loss": 3.8655,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558239095156264e-05,
+      "loss": 3.8729,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557400500405212e-05,
+      "loss": 3.8736,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55656190565416e-05,
+      "loss": 3.8817,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555724948783481e-05,
+      "loss": 3.8689,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554886354032429e-05,
+      "loss": 3.8562,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554047759281377e-05,
+      "loss": 3.8662,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553209164530325e-05,
+      "loss": 3.8661,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552372207659646e-05,
+      "loss": 3.8748,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551533612908594e-05,
+      "loss": 3.8729,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550695018157542e-05,
+      "loss": 3.8679,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54985642340649e-05,
+      "loss": 3.8725,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549019466535811e-05,
+      "loss": 3.8576,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548180871784759e-05,
+      "loss": 3.8585,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547342277033707e-05,
+      "loss": 3.8639,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546503682282655e-05,
+      "loss": 3.8682,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545666725411976e-05,
+      "loss": 3.87,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544828130660924e-05,
+      "loss": 3.8748,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543989535909872e-05,
+      "loss": 3.8485,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54315094115882e-05,
+      "loss": 3.8721,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542313984288141e-05,
+      "loss": 3.8751,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541475389537089e-05,
+      "loss": 3.8662,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540636794786037e-05,
+      "loss": 3.8639,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539798200034985e-05,
+      "loss": 3.8535,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389612431643066e-05,
+      "loss": 3.8624,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381226484132546e-05,
+      "loss": 3.8659,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372840536622026e-05,
+      "loss": 3.8471,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364454589111506e-05,
+      "loss": 3.8576,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356085020404715e-05,
+      "loss": 3.8596,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347699072894195e-05,
+      "loss": 3.8645,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339329504187404e-05,
+      "loss": 3.8465,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330943556676884e-05,
+      "loss": 3.8689,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322557609166364e-05,
+      "loss": 3.8441,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314171661655844e-05,
+      "loss": 3.8691,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305785714145324e-05,
+      "loss": 3.8569,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297399766634804e-05,
+      "loss": 3.8504,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5289013819124284e-05,
+      "loss": 3.8678,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528062787161377e-05,
+      "loss": 3.8674,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527227468171071e-05,
+      "loss": 3.8596,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526388873420019e-05,
+      "loss": 3.8537,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525550278668967e-05,
+      "loss": 3.8578,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524711683917915e-05,
+      "loss": 3.858,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523873089166863e-05,
+      "loss": 3.8481,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523034494415811e-05,
+      "loss": 3.8556,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522195899664759e-05,
+      "loss": 3.8543,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521357304913707e-05,
+      "loss": 3.8591,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520520348043028e-05,
+      "loss": 3.8548,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519681753291976e-05,
+      "loss": 3.8547,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518843158540924e-05,
+      "loss": 3.8389,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5180045637898725e-05,
+      "loss": 3.8563,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171676069191934e-05,
+      "loss": 3.8482,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163290121681414e-05,
+      "loss": 3.8629,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154904174170894e-05,
+      "loss": 3.861,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146518226660373e-05,
+      "loss": 3.8567,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513814865795358e-05,
+      "loss": 3.8406,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512976271044306e-05,
+      "loss": 3.8601,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512137676293254e-05,
+      "loss": 3.8554,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511299081542202e-05,
+      "loss": 3.8635,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510462124671523e-05,
+      "loss": 3.8432,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509623529920471e-05,
+      "loss": 3.859,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508786573049793e-05,
+      "loss": 3.8613,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507947978298741e-05,
+      "loss": 3.8669,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507109383547689e-05,
+      "loss": 3.8436,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506270788796637e-05,
+      "loss": 3.8554,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505432194045585e-05,
+      "loss": 3.853,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504593599294533e-05,
+      "loss": 3.8435,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50375500454348e-05,
+      "loss": 3.8548,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502916409792428e-05,
+      "loss": 3.8591,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020794529217496e-05,
+      "loss": 3.8492,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012408581706976e-05,
+      "loss": 3.8468,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500402263419645e-05,
+      "loss": 3.8486,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.910466432571411,
+      "eval_runtime": 306.1837,
+      "eval_samples_per_second": 1246.281,
+      "eval_steps_per_second": 38.947,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499563668668593e-05,
+      "loss": 3.8283,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987250739175416e-05,
+      "loss": 3.8344,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978864791664896e-05,
+      "loss": 3.8528,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970478844154376e-05,
+      "loss": 3.8461,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962092896643856e-05,
+      "loss": 3.8591,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953706949133336e-05,
+      "loss": 3.8393,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945321001622816e-05,
+      "loss": 3.8444,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936935054112296e-05,
+      "loss": 3.8264,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928565485405505e-05,
+      "loss": 3.8443,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920179537894985e-05,
+      "loss": 3.8434,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911793590384465e-05,
+      "loss": 3.8437,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903407642873945e-05,
+      "loss": 3.8495,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489505445297088e-05,
+      "loss": 3.8354,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488666850546037e-05,
+      "loss": 3.8364,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487828255794985e-05,
+      "loss": 3.8301,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486989661043933e-05,
+      "loss": 3.8251,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486151066292881e-05,
+      "loss": 3.8343,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485312471541829e-05,
+      "loss": 3.8324,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484473876790777e-05,
+      "loss": 3.8303,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483635282039725e-05,
+      "loss": 3.8524,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482798325169046e-05,
+      "loss": 3.836,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481959730417994e-05,
+      "loss": 3.8348,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481121135666942e-05,
+      "loss": 3.8436,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48028254091589e-05,
+      "loss": 3.8455,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479445584045211e-05,
+      "loss": 3.8278,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478606989294159e-05,
+      "loss": 3.8339,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477768394543107e-05,
+      "loss": 3.8376,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476929799792055e-05,
+      "loss": 3.8344,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476092842921376e-05,
+      "loss": 3.8199,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475254248170324e-05,
+      "loss": 3.8202,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474415653419272e-05,
+      "loss": 3.8384,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47357705866822e-05,
+      "loss": 3.8404,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472740101797541e-05,
+      "loss": 3.835,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471901507046489e-05,
+      "loss": 3.8362,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471062912295437e-05,
+      "loss": 3.8278,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470224317544385e-05,
+      "loss": 3.8353,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469387360673706e-05,
+      "loss": 3.8211,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468548765922654e-05,
+      "loss": 3.8308,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467710171171602e-05,
+      "loss": 3.8126,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46687157642055e-05,
+      "loss": 3.8239,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466034619549872e-05,
+      "loss": 3.8215,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46519602479882e-05,
+      "loss": 3.8404,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464357430047768e-05,
+      "loss": 3.8284,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463518835296716e-05,
+      "loss": 3.8254,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626818784260366e-05,
+      "loss": 3.8289,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618432836749846e-05,
+      "loss": 3.8306,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4610046889239326e-05,
+      "loss": 3.8328,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601660941728806e-05,
+      "loss": 3.8116,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593291373022015e-05,
+      "loss": 3.8186,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584905425511495e-05,
+      "loss": 3.8333,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576519478000975e-05,
+      "loss": 3.8287,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4568133530490455e-05,
+      "loss": 3.8184,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455976396178367e-05,
+      "loss": 3.8137,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455137801427315e-05,
+      "loss": 3.8203,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454299206676263e-05,
+      "loss": 3.8082,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453460611925211e-05,
+      "loss": 3.8272,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452623655054532e-05,
+      "loss": 3.8114,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45178506030348e-05,
+      "loss": 3.8217,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450946465552428e-05,
+      "loss": 3.8205,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450107870801376e-05,
+      "loss": 3.8144,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449270913930697e-05,
+      "loss": 3.8097,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448432319179645e-05,
+      "loss": 3.8137,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447593724428593e-05,
+      "loss": 3.8058,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446755129677541e-05,
+      "loss": 3.8137,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4459181728068624e-05,
+      "loss": 3.8218,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450795780558104e-05,
+      "loss": 3.8097,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442409833047584e-05,
+      "loss": 3.8099,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443402388553706e-05,
+      "loss": 3.7979,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442565431683027e-05,
+      "loss": 3.8047,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441726836931975e-05,
+      "loss": 3.8184,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440888242180923e-05,
+      "loss": 3.8248,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400496474298706e-05,
+      "loss": 3.8124,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439212690559192e-05,
+      "loss": 3.8211,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43837409580814e-05,
+      "loss": 3.8261,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437535501057088e-05,
+      "loss": 3.8205,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436696906306036e-05,
+      "loss": 3.8003,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435859949435358e-05,
+      "loss": 3.8179,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435021354684306e-05,
+      "loss": 3.8118,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434182759933253e-05,
+      "loss": 3.8049,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433344165182201e-05,
+      "loss": 3.8089,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432507208311523e-05,
+      "loss": 3.8155,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431668613560471e-05,
+      "loss": 3.8236,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430830018809418e-05,
+      "loss": 3.809,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429991424058366e-05,
+      "loss": 3.7927,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291544671876876e-05,
+      "loss": 3.8067,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283158724366356e-05,
+      "loss": 3.8079,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274772776855836e-05,
+      "loss": 3.8104,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266386829345316e-05,
+      "loss": 3.8123,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425801726063853e-05,
+      "loss": 3.8103,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249631313128005e-05,
+      "loss": 3.8127,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241245365617485e-05,
+      "loss": 3.8018,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42328757969107e-05,
+      "loss": 3.7974,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422448984940018e-05,
+      "loss": 3.804,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216103901889654e-05,
+      "loss": 3.8125,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207717954379134e-05,
+      "loss": 3.8114,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419934838567235e-05,
+      "loss": 3.8158,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419096243816182e-05,
+      "loss": 3.7947,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418257649065131e-05,
+      "loss": 3.809,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417419054314079e-05,
+      "loss": 3.814,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165820974434005e-05,
+      "loss": 3.8112,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415743502692348e-05,
+      "loss": 3.8046,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414904907941296e-05,
+      "loss": 3.8003,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414066313190244e-05,
+      "loss": 3.8029,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132293563195654e-05,
+      "loss": 3.8097,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412390761568513e-05,
+      "loss": 3.7865,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411552166817461e-05,
+      "loss": 3.802,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410713572066409e-05,
+      "loss": 3.8008,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098766151957296e-05,
+      "loss": 3.8062,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409039658325051e-05,
+      "loss": 3.7896,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408201063573999e-05,
+      "loss": 3.8083,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407362468822948e-05,
+      "loss": 3.7975,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406523874071895e-05,
+      "loss": 3.8071,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405685279320843e-05,
+      "loss": 3.7988,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404846684569791e-05,
+      "loss": 3.7938,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404008089818739e-05,
+      "loss": 3.8098,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403169495067687e-05,
+      "loss": 3.8127,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402334176077382e-05,
+      "loss": 3.8037,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40149558132633e-05,
+      "loss": 3.7947,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400656986575277e-05,
+      "loss": 3.8024,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399818391824225e-05,
+      "loss": 3.8028,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398979797073173e-05,
+      "loss": 3.7901,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398141202322122e-05,
+      "loss": 3.7998,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39730260757107e-05,
+      "loss": 3.7969,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396464012820018e-05,
+      "loss": 3.8014,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956270559493386e-05,
+      "loss": 3.7969,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947884611982866e-05,
+      "loss": 3.8012,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939498664472346e-05,
+      "loss": 3.7854,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3931112716961826e-05,
+      "loss": 3.7992,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922743148255035e-05,
+      "loss": 3.7951,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914357200744515e-05,
+      "loss": 3.8072,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905971253233995e-05,
+      "loss": 3.7977,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897585305723475e-05,
+      "loss": 3.8051,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889215737016684e-05,
+      "loss": 3.7865,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388082978950617e-05,
+      "loss": 3.8059,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387244384199565e-05,
+      "loss": 3.8026,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386405789448513e-05,
+      "loss": 3.8058,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385568832577834e-05,
+      "loss": 3.7879,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384730237826782e-05,
+      "loss": 3.8044,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383893280956103e-05,
+      "loss": 3.8032,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383054686205051e-05,
+      "loss": 3.8174,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382216091453999e-05,
+      "loss": 3.7885,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381377496702947e-05,
+      "loss": 3.7999,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380538901951895e-05,
+      "loss": 3.7974,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379700307200843e-05,
+      "loss": 3.7883,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378861712449791e-05,
+      "loss": 3.8023,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378023117698739e-05,
+      "loss": 3.8038,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771861608280604e-05,
+      "loss": 3.7981,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763475660770084e-05,
+      "loss": 3.7889,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3755089713259564e-05,
+      "loss": 3.7978,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.885854959487915,
+      "eval_runtime": 305.8448,
+      "eval_samples_per_second": 1247.662,
+      "eval_steps_per_second": 38.99,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3746703765749044e-05,
+      "loss": 3.7852,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373833419704225e-05,
+      "loss": 3.7776,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372994824953173e-05,
+      "loss": 3.7987,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372156230202121e-05,
+      "loss": 3.794,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371317635451069e-05,
+      "loss": 3.7999,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37048067858039e-05,
+      "loss": 3.7891,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369642083829338e-05,
+      "loss": 3.7874,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368803489078286e-05,
+      "loss": 3.778,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367964894327234e-05,
+      "loss": 3.7902,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367127937456556e-05,
+      "loss": 3.7908,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366289342705504e-05,
+      "loss": 3.7903,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365450747954452e-05,
+      "loss": 3.7963,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364613791083773e-05,
+      "loss": 3.7849,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363775196332721e-05,
+      "loss": 3.7844,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3629366015816687e-05,
+      "loss": 3.7809,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620980068306167e-05,
+      "loss": 3.7724,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612594120795647e-05,
+      "loss": 3.7842,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3604208173285126e-05,
+      "loss": 3.7793,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595822225774606e-05,
+      "loss": 3.7751,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358743627826408e-05,
+      "loss": 3.7971,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3579066709557295e-05,
+      "loss": 3.7883,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3570680762046775e-05,
+      "loss": 3.7872,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3562294814536255e-05,
+      "loss": 3.791,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3553908867025735e-05,
+      "loss": 3.7955,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354553929831895e-05,
+      "loss": 3.7718,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353715335080843e-05,
+      "loss": 3.7831,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528767403297904e-05,
+      "loss": 3.7893,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352039783459112e-05,
+      "loss": 3.7785,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35120118870806e-05,
+      "loss": 3.7714,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350362593957008e-05,
+      "loss": 3.7687,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349523999205955e-05,
+      "loss": 3.7878,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348687042335277e-05,
+      "loss": 3.788,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347848447584225e-05,
+      "loss": 3.7847,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347009852833173e-05,
+      "loss": 3.7876,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346171258082121e-05,
+      "loss": 3.778,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345332663331069e-05,
+      "loss": 3.7821,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344494068580017e-05,
+      "loss": 3.774,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343655473828965e-05,
+      "loss": 3.7811,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342816879077913e-05,
+      "loss": 3.7641,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341979922207234e-05,
+      "loss": 3.771,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341141327456182e-05,
+      "loss": 3.7723,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34030273270513e-05,
+      "loss": 3.7897,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339465775834451e-05,
+      "loss": 3.7798,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338627181083399e-05,
+      "loss": 3.7754,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337788586332347e-05,
+      "loss": 3.7831,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336949991581295e-05,
+      "loss": 3.7766,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336113034710616e-05,
+      "loss": 3.7831,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335274439959564e-05,
+      "loss": 3.767,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334435845208512e-05,
+      "loss": 3.7654,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33359725045746e-05,
+      "loss": 3.782,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332760293586781e-05,
+      "loss": 3.779,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331921698835729e-05,
+      "loss": 3.7684,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331083104084677e-05,
+      "loss": 3.7687,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330244509333625e-05,
+      "loss": 3.7715,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329407552462946e-05,
+      "loss": 3.7556,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328568957711894e-05,
+      "loss": 3.7798,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327730362960842e-05,
+      "loss": 3.7651,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32689176820979e-05,
+      "loss": 3.7694,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260548113391116e-05,
+      "loss": 3.7768,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3252162165880596e-05,
+      "loss": 3.7658,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243776218370076e-05,
+      "loss": 3.7635,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3235390270859556e-05,
+      "loss": 3.7659,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3227020702152765e-05,
+      "loss": 3.7595,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218634754642245e-05,
+      "loss": 3.7622,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3210248807131725e-05,
+      "loss": 3.7743,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201862859621205e-05,
+      "loss": 3.7659,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193493290914414e-05,
+      "loss": 3.7593,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185107343403894e-05,
+      "loss": 3.7547,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176721395893374e-05,
+      "loss": 3.7545,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168335448382854e-05,
+      "loss": 3.7656,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315996587967607e-05,
+      "loss": 3.7754,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315157993216555e-05,
+      "loss": 3.7675,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314319398465503e-05,
+      "loss": 3.7738,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313480803714451e-05,
+      "loss": 3.7763,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312643846843772e-05,
+      "loss": 3.7791,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31180525209272e-05,
+      "loss": 3.7518,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310966657341668e-05,
+      "loss": 3.7688,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310128062590616e-05,
+      "loss": 3.7679,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309291105719937e-05,
+      "loss": 3.7534,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308452510968885e-05,
+      "loss": 3.7671,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307613916217833e-05,
+      "loss": 3.7654,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306775321466781e-05,
+      "loss": 3.7754,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3059383645961024e-05,
+      "loss": 3.7631,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050997698450504e-05,
+      "loss": 3.7469,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042611750939984e-05,
+      "loss": 3.7602,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034225803429464e-05,
+      "loss": 3.761,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302585623472267e-05,
+      "loss": 3.7676,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301747028721215e-05,
+      "loss": 3.7617,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300908433970163e-05,
+      "loss": 3.7666,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300069839219111e-05,
+      "loss": 3.765,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299232882348432e-05,
+      "loss": 3.7573,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29839428759738e-05,
+      "loss": 3.7486,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297555692846328e-05,
+      "loss": 3.7584,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296717098095276e-05,
+      "loss": 3.7684,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295880141224598e-05,
+      "loss": 3.7635,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295041546473546e-05,
+      "loss": 3.7675,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294202951722494e-05,
+      "loss": 3.7521,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293364356971442e-05,
+      "loss": 3.7635,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2925274001007626e-05,
+      "loss": 3.7675,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916888053497106e-05,
+      "loss": 3.767,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908502105986586e-05,
+      "loss": 3.7575,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2900116158476066e-05,
+      "loss": 3.7569,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2891746589769275e-05,
+      "loss": 3.7571,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2883360642258755e-05,
+      "loss": 3.7641,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2874974694748235e-05,
+      "loss": 3.74,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286660512604145e-05,
+      "loss": 3.757,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285821917853093e-05,
+      "loss": 3.7506,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284983323102041e-05,
+      "loss": 3.764,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284146366231362e-05,
+      "loss": 3.7444,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28330777148031e-05,
+      "loss": 3.769,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282469176729258e-05,
+      "loss": 3.7516,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281630581978206e-05,
+      "loss": 3.7626,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280791987227154e-05,
+      "loss": 3.753,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279953392476102e-05,
+      "loss": 3.7513,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27911479772505e-05,
+      "loss": 3.7609,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278276202973998e-05,
+      "loss": 3.7698,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277440883983692e-05,
+      "loss": 3.759,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2766022892326405e-05,
+      "loss": 3.7482,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757636944815885e-05,
+      "loss": 3.757,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749250997305365e-05,
+      "loss": 3.7595,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740865049794845e-05,
+      "loss": 3.748,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2732479102284325e-05,
+      "loss": 3.7581,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2724093154773805e-05,
+      "loss": 3.7473,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715707207263285e-05,
+      "loss": 3.7558,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2707337638556494e-05,
+      "loss": 3.7562,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698951691045974e-05,
+      "loss": 3.7605,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690565743535454e-05,
+      "loss": 3.7379,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268217979602493e-05,
+      "loss": 3.7559,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267381022731814e-05,
+      "loss": 3.7494,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266542427980762e-05,
+      "loss": 3.7635,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26570383322971e-05,
+      "loss": 3.7561,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264865238478658e-05,
+      "loss": 3.7612,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26402828160798e-05,
+      "loss": 3.7413,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263189686856928e-05,
+      "loss": 3.7613,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262351092105876e-05,
+      "loss": 3.7594,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261512497354823e-05,
+      "loss": 3.7599,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260675540484145e-05,
+      "loss": 3.7491,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259836945733093e-05,
+      "loss": 3.7582,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25899835098204e-05,
+      "loss": 3.7615,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258159756230988e-05,
+      "loss": 3.7707,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2573244372406825e-05,
+      "loss": 3.7467,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256485842489631e-05,
+      "loss": 3.758,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255647247738579e-05,
+      "loss": 3.7551,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254808652987527e-05,
+      "loss": 3.745,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253970058236475e-05,
+      "loss": 3.7602,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253131463485423e-05,
+      "loss": 3.7565,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522928687343705e-05,
+      "loss": 3.7612,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514542739833185e-05,
+      "loss": 3.7447,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25061731711264e-05,
+      "loss": 3.7536,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.869858503341675,
+      "eval_runtime": 310.6177,
+      "eval_samples_per_second": 1228.491,
+      "eval_steps_per_second": 38.391,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497787223615874e-05,
+      "loss": 3.7439,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2489401276105354e-05,
+      "loss": 3.7306,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2481015328594834e-05,
+      "loss": 3.7564,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472629381084314e-05,
+      "loss": 3.7523,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464243433573794e-05,
+      "loss": 3.7592,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245585748606328e-05,
+      "loss": 3.7486,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244747153855276e-05,
+      "loss": 3.7392,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243908559104224e-05,
+      "loss": 3.7394,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243071602233545e-05,
+      "loss": 3.7465,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242233007482493e-05,
+      "loss": 3.747,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241394412731441e-05,
+      "loss": 3.7503,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240555817980389e-05,
+      "loss": 3.755,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239720498990083e-05,
+      "loss": 3.7445,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238881904239031e-05,
+      "loss": 3.7389,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238043309487979e-05,
+      "loss": 3.7429,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237204714736927e-05,
+      "loss": 3.7283,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236366119985875e-05,
+      "loss": 3.7413,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2355275252348234e-05,
+      "loss": 3.7392,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346889304837714e-05,
+      "loss": 3.7354,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2338503357327194e-05,
+      "loss": 3.7565,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23301337886204e-05,
+      "loss": 3.747,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232174784110988e-05,
+      "loss": 3.7468,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231336189359936e-05,
+      "loss": 3.7472,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230499232489257e-05,
+      "loss": 3.7524,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229660637738205e-05,
+      "loss": 3.7319,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228822042987153e-05,
+      "loss": 3.7452,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227983448236101e-05,
+      "loss": 3.7425,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227144853485049e-05,
+      "loss": 3.7374,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226306258733997e-05,
+      "loss": 3.7336,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225467663982945e-05,
+      "loss": 3.7251,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224629069231893e-05,
+      "loss": 3.7491,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223793750241588e-05,
+      "loss": 3.7462,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222955155490536e-05,
+      "loss": 3.7434,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222116560739484e-05,
+      "loss": 3.7474,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221277965988432e-05,
+      "loss": 3.7364,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22043937123738e-05,
+      "loss": 3.7417,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219600776486328e-05,
+      "loss": 3.7363,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218762181735276e-05,
+      "loss": 3.7347,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2179252248645966e-05,
+      "loss": 3.7252,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170866301135446e-05,
+      "loss": 3.7309,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2162480353624926e-05,
+      "loss": 3.7338,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2154094406114406e-05,
+      "loss": 3.7494,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214572483740762e-05,
+      "loss": 3.7361,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21373388898971e-05,
+      "loss": 3.7338,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212895294238658e-05,
+      "loss": 3.7461,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212056699487606e-05,
+      "loss": 3.7354,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211219742616927e-05,
+      "loss": 3.7432,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210381147865875e-05,
+      "loss": 3.7285,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209542553114823e-05,
+      "loss": 3.7205,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208703958363771e-05,
+      "loss": 3.742,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078653636127184e-05,
+      "loss": 3.7387,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070267688616664e-05,
+      "loss": 3.7338,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206189811990988e-05,
+      "loss": 3.7264,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205351217239936e-05,
+      "loss": 3.7335,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204512622488884e-05,
+      "loss": 3.7128,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203674027737832e-05,
+      "loss": 3.7378,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20283543298678e-05,
+      "loss": 3.728,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201996838235728e-05,
+      "loss": 3.7266,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201158243484676e-05,
+      "loss": 3.7391,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2003229244943704e-05,
+      "loss": 3.7252,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994843297433184e-05,
+      "loss": 3.7223,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198645734992266e-05,
+      "loss": 3.7302,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197807140241214e-05,
+      "loss": 3.718,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196968545490162e-05,
+      "loss": 3.725,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196131588619483e-05,
+      "loss": 3.7327,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195292993868431e-05,
+      "loss": 3.7272,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194454399117379e-05,
+      "loss": 3.72,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193615804366327e-05,
+      "loss": 3.7174,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192777209615275e-05,
+      "loss": 3.7126,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191938614864223e-05,
+      "loss": 3.7254,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191100020113171e-05,
+      "loss": 3.7372,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190263063242492e-05,
+      "loss": 3.728,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18942446849144e-05,
+      "loss": 3.7366,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188587511620761e-05,
+      "loss": 3.7338,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187748916869709e-05,
+      "loss": 3.7414,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186910322118657e-05,
+      "loss": 3.7167,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186071727367605e-05,
+      "loss": 3.7329,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185233132616553e-05,
+      "loss": 3.7299,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184394537865501e-05,
+      "loss": 3.7118,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18355594311445e-05,
+      "loss": 3.7306,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182718986243771e-05,
+      "loss": 3.7297,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818803914927187e-05,
+      "loss": 3.7348,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1810417967416667e-05,
+      "loss": 3.7259,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1802032019906146e-05,
+      "loss": 3.7118,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793646072395626e-05,
+      "loss": 3.719,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1785260124885106e-05,
+      "loss": 3.722,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776874177374586e-05,
+      "loss": 3.7285,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1768504608667795e-05,
+      "loss": 3.7229,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760118661157275e-05,
+      "loss": 3.7302,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751732713646755e-05,
+      "loss": 3.7268,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1743346766136235e-05,
+      "loss": 3.722,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173497719742945e-05,
+      "loss": 3.7113,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172659124991893e-05,
+      "loss": 3.7195,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171820530240841e-05,
+      "loss": 3.7327,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170981935489789e-05,
+      "loss": 3.7289,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17014497861911e-05,
+      "loss": 3.7276,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169306383868058e-05,
+      "loss": 3.7135,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168467789117006e-05,
+      "loss": 3.7237,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167629194365954e-05,
+      "loss": 3.7303,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166792237495275e-05,
+      "loss": 3.7269,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165955280624596e-05,
+      "loss": 3.7224,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165116685873544e-05,
+      "loss": 3.7204,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164278091122492e-05,
+      "loss": 3.7214,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1634394963714405e-05,
+      "loss": 3.7256,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1626009016203885e-05,
+      "loss": 3.7047,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617623068693365e-05,
+      "loss": 3.7155,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609237121182845e-05,
+      "loss": 3.7148,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160085117367232e-05,
+      "loss": 3.7276,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159249798376926e-05,
+      "loss": 3.7098,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158411203625874e-05,
+      "loss": 3.7275,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157572608874822e-05,
+      "loss": 3.7158,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15673401412377e-05,
+      "loss": 3.7224,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155895419372718e-05,
+      "loss": 3.7226,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155056824621666e-05,
+      "loss": 3.7118,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154218229870614e-05,
+      "loss": 3.7234,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153379635119562e-05,
+      "loss": 3.7327,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152544316129257e-05,
+      "loss": 3.7207,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151705721378205e-05,
+      "loss": 3.719,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150867126627153e-05,
+      "loss": 3.7119,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150028531876101e-05,
+      "loss": 3.7243,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149189937125049e-05,
+      "loss": 3.7166,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148351342373997e-05,
+      "loss": 3.7166,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1475143855033176e-05,
+      "loss": 3.709,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466757907522656e-05,
+      "loss": 3.7189,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1458371960012136e-05,
+      "loss": 3.7203,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144998601250161e-05,
+      "loss": 3.7221,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1441600064991096e-05,
+      "loss": 3.7031,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1433214117480576e-05,
+      "loss": 3.7227,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424828169970056e-05,
+      "loss": 3.7104,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1416458601263265e-05,
+      "loss": 3.7291,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1408072653752745e-05,
+      "loss": 3.7173,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399686706242225e-05,
+      "loss": 3.728,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1391300758731705e-05,
+      "loss": 3.7051,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382914811221185e-05,
+      "loss": 3.7273,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1374528863710665e-05,
+      "loss": 3.7217,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1366159295003874e-05,
+      "loss": 3.7252,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357773347493354e-05,
+      "loss": 3.7077,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349387399982834e-05,
+      "loss": 3.7221,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1341001452472314e-05,
+      "loss": 3.7259,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332615504961794e-05,
+      "loss": 3.7334,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132424593625501e-05,
+      "loss": 3.7132,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131585998874449e-05,
+      "loss": 3.7211,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130747404123397e-05,
+      "loss": 3.7221,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129908809372345e-05,
+      "loss": 3.7138,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129070214621293e-05,
+      "loss": 3.7203,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128233257750614e-05,
+      "loss": 3.7211,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127394662999562e-05,
+      "loss": 3.7223,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12655606824851e-05,
+      "loss": 3.7118,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125717473497458e-05,
+      "loss": 3.7183,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.859259843826294,
+      "eval_runtime": 306.5299,
+      "eval_samples_per_second": 1244.874,
+      "eval_steps_per_second": 38.903,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124878878746406e-05,
+      "loss": 3.7126,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124040283995354e-05,
+      "loss": 3.6957,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123201689244302e-05,
+      "loss": 3.7179,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122364732373623e-05,
+      "loss": 3.7183,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1215261376225714e-05,
+      "loss": 3.7196,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206875428715194e-05,
+      "loss": 3.7141,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1198489481204674e-05,
+      "loss": 3.707,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1190103533694154e-05,
+      "loss": 3.7036,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118171758618363e-05,
+      "loss": 3.7119,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117333163867311e-05,
+      "loss": 3.7091,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116496206996632e-05,
+      "loss": 3.7142,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156576122455796e-05,
+      "loss": 3.7196,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114820655374901e-05,
+      "loss": 3.7118,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113982060623849e-05,
+      "loss": 3.7059,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113143465872797e-05,
+      "loss": 3.7046,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112304871121745e-05,
+      "loss": 3.6904,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111466276370693e-05,
+      "loss": 3.7073,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110627681619641e-05,
+      "loss": 3.7053,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109789086868589e-05,
+      "loss": 3.7039,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108950492117537e-05,
+      "loss": 3.7166,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108111897366485e-05,
+      "loss": 3.7166,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107273302615433e-05,
+      "loss": 3.7103,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106434707864381e-05,
+      "loss": 3.712,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105597750993702e-05,
+      "loss": 3.715,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10475915624265e-05,
+      "loss": 3.6996,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103920561491598e-05,
+      "loss": 3.7115,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103081966740546e-05,
+      "loss": 3.7107,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102243371989494e-05,
+      "loss": 3.7039,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101406415118815e-05,
+      "loss": 3.6979,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005678203677637e-05,
+      "loss": 3.6964,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997292256167117e-05,
+      "loss": 3.7098,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988906308656597e-05,
+      "loss": 3.7151,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980520361146076e-05,
+      "loss": 3.7102,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972150792439286e-05,
+      "loss": 3.7121,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963764844928766e-05,
+      "loss": 3.6978,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955378897418245e-05,
+      "loss": 3.709,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0946992949907725e-05,
+      "loss": 3.7089,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938623381200935e-05,
+      "loss": 3.6966,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0930237433690414e-05,
+      "loss": 3.6966,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921851486179894e-05,
+      "loss": 3.6954,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0913465538669374e-05,
+      "loss": 3.6981,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0905079591158854e-05,
+      "loss": 3.7162,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896693643648334e-05,
+      "loss": 3.703,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088832407494155e-05,
+      "loss": 3.6969,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087993812743103e-05,
+      "loss": 3.7115,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087155217992051e-05,
+      "loss": 3.7044,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086316623240999e-05,
+      "loss": 3.7073,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085478028489946e-05,
+      "loss": 3.6947,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084639433738894e-05,
+      "loss": 3.6859,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083800838987842e-05,
+      "loss": 3.7138,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08296224423679e-05,
+      "loss": 3.7018,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082125287366111e-05,
+      "loss": 3.6998,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081288330495433e-05,
+      "loss": 3.6976,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080449735744381e-05,
+      "loss": 3.6917,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079611140993329e-05,
+      "loss": 3.6861,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078772546242277e-05,
+      "loss": 3.7008,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077933951491225e-05,
+      "loss": 3.6965,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770969946205464e-05,
+      "loss": 3.6972,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076258399869494e-05,
+      "loss": 3.7041,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075419805118442e-05,
+      "loss": 3.6934,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07458121036739e-05,
+      "loss": 3.6888,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0737442534967106e-05,
+      "loss": 3.7006,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0729056587456586e-05,
+      "loss": 3.6828,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0720670639946066e-05,
+      "loss": 3.6918,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0712284692435546e-05,
+      "loss": 3.6988,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0703898744925026e-05,
+      "loss": 3.6958,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0695512797414506e-05,
+      "loss": 3.687,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068712684990399e-05,
+      "loss": 3.6865,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067874090239347e-05,
+      "loss": 3.6772,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067037133368668e-05,
+      "loss": 3.6946,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066198538617616e-05,
+      "loss": 3.7027,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065361581746937e-05,
+      "loss": 3.6977,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064522986995885e-05,
+      "loss": 3.699,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063684392244833e-05,
+      "loss": 3.7016,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062845797493781e-05,
+      "loss": 3.7108,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062007202742729e-05,
+      "loss": 3.6813,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061168607991677e-05,
+      "loss": 3.7022,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060330013240625e-05,
+      "loss": 3.6981,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059493056369946e-05,
+      "loss": 3.675,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0586544616188946e-05,
+      "loss": 3.7027,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0578158668678426e-05,
+      "loss": 3.6918,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0569772721167906e-05,
+      "loss": 3.7041,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0561403152461115e-05,
+      "loss": 3.6963,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0553017204950595e-05,
+      "loss": 3.6793,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0544631257440075e-05,
+      "loss": 3.6842,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0536245309929555e-05,
+      "loss": 3.6931,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0527859362419035e-05,
+      "loss": 3.694,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0519473414908515e-05,
+      "loss": 3.6917,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0511087467397995e-05,
+      "loss": 3.6978,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050270151988747e-05,
+      "loss": 3.6924,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0494331951180684e-05,
+      "loss": 3.6889,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0485946003670164e-05,
+      "loss": 3.686,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477560056159644e-05,
+      "loss": 3.6839,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0469174108649124e-05,
+      "loss": 3.6997,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046080453994234e-05,
+      "loss": 3.6959,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045243497123555e-05,
+      "loss": 3.701,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044404902372503e-05,
+      "loss": 3.6785,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043566307621451e-05,
+      "loss": 3.6925,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042727712870399e-05,
+      "loss": 3.6975,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041889118119347e-05,
+      "loss": 3.697,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041052161248668e-05,
+      "loss": 3.6904,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040213566497616e-05,
+      "loss": 3.6863,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039374971746564e-05,
+      "loss": 3.6897,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038536376995512e-05,
+      "loss": 3.6932,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03769778224446e-05,
+      "loss": 3.6721,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036859187493408e-05,
+      "loss": 3.6899,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036020592742356e-05,
+      "loss": 3.6796,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351836358716766e-05,
+      "loss": 3.6983,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0343450411206246e-05,
+      "loss": 3.6795,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0335064463695726e-05,
+      "loss": 3.6906,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326678516185206e-05,
+      "loss": 3.6876,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0318292568674686e-05,
+      "loss": 3.695,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309906621164166e-05,
+      "loss": 3.6891,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0301537052457375e-05,
+      "loss": 3.6787,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0293151104946855e-05,
+      "loss": 3.6915,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284765157436335e-05,
+      "loss": 3.6995,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0276379209925815e-05,
+      "loss": 3.6897,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02679932624153e-05,
+      "loss": 3.6903,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025960731490478e-05,
+      "loss": 3.679,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025123774619799e-05,
+      "loss": 3.6909,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024285179868747e-05,
+      "loss": 3.6903,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023446585117695e-05,
+      "loss": 3.6808,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022607990366643e-05,
+      "loss": 3.684,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021769395615591e-05,
+      "loss": 3.6864,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020930800864539e-05,
+      "loss": 3.6911,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020092206113487e-05,
+      "loss": 3.6924,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019255249242808e-05,
+      "loss": 3.6728,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018416654491756e-05,
+      "loss": 3.6922,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017578059740704e-05,
+      "loss": 3.6767,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0167411028700256e-05,
+      "loss": 3.6966,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0159025081189736e-05,
+      "loss": 3.6875,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0150639133679216e-05,
+      "loss": 3.6963,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0142253186168696e-05,
+      "loss": 3.6744,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0133867238658175e-05,
+      "loss": 3.6982,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0125497669951385e-05,
+      "loss": 3.6894,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0117111722440865e-05,
+      "loss": 3.6931,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108725774930344e-05,
+      "loss": 3.6822,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0100339827419824e-05,
+      "loss": 3.6861,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0091953879909304e-05,
+      "loss": 3.6916,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008356793239878e-05,
+      "loss": 3.7067,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0075198363691993e-05,
+      "loss": 3.6831,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006681241618147e-05,
+      "loss": 3.6892,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005842646867095e-05,
+      "loss": 3.6953,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005004052116043e-05,
+      "loss": 3.6821,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004167095245365e-05,
+      "loss": 3.6837,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003328500494313e-05,
+      "loss": 3.6919,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00248990574326e-05,
+      "loss": 3.6921,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001651310992208e-05,
+      "loss": 3.6824,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000812716241156e-05,
+      "loss": 3.6884,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8526220321655273,
+      "eval_runtime": 352.6196,
+      "eval_samples_per_second": 1082.16,
+      "eval_steps_per_second": 33.818,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999974121490104e-05,
+      "loss": 3.6745,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999135526739052e-05,
+      "loss": 3.6697,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998296931988e-05,
+      "loss": 3.6835,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997458337236948e-05,
+      "loss": 3.686,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996619742485896e-05,
+      "loss": 3.6933,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995782785615218e-05,
+      "loss": 3.6805,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994944190864166e-05,
+      "loss": 3.6784,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994105596113114e-05,
+      "loss": 3.6751,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993267001362062e-05,
+      "loss": 3.6784,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99242840661101e-05,
+      "loss": 3.6813,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991589811859958e-05,
+      "loss": 3.6842,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990752854989279e-05,
+      "loss": 3.686,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9899158981185996e-05,
+      "loss": 3.6819,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890773033675476e-05,
+      "loss": 3.6789,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882387086164956e-05,
+      "loss": 3.6772,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9874001138654436e-05,
+      "loss": 3.661,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865615191143916e-05,
+      "loss": 3.6742,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857229243633396e-05,
+      "loss": 3.6758,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848843296122876e-05,
+      "loss": 3.6758,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9840457348612356e-05,
+      "loss": 3.6812,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983208777990557e-05,
+      "loss": 3.6879,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982370183239505e-05,
+      "loss": 3.6815,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981531588488453e-05,
+      "loss": 3.6822,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980692993737401e-05,
+      "loss": 3.6842,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979854398986349e-05,
+      "loss": 3.6726,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790158042352964e-05,
+      "loss": 3.6822,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781772094842444e-05,
+      "loss": 3.6806,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773386147331924e-05,
+      "loss": 3.6751,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976503295742887e-05,
+      "loss": 3.6728,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975664700991835e-05,
+      "loss": 3.6648,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974826106240783e-05,
+      "loss": 3.6784,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973987511489731e-05,
+      "loss": 3.6878,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973148916738679e-05,
+      "loss": 3.6811,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9723119598680005e-05,
+      "loss": 3.6777,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9714733651169485e-05,
+      "loss": 3.6685,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9706347703658965e-05,
+      "loss": 3.6823,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969796175614844e-05,
+      "loss": 3.6775,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689592187441654e-05,
+      "loss": 3.6715,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9681206239931134e-05,
+      "loss": 3.6679,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672820292420614e-05,
+      "loss": 3.6637,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966443434491009e-05,
+      "loss": 3.6652,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96560647762033e-05,
+      "loss": 3.6868,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964767882869278e-05,
+      "loss": 3.6737,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963929288118226e-05,
+      "loss": 3.6706,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963090693367174e-05,
+      "loss": 3.6806,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962252098616122e-05,
+      "loss": 3.6736,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96141350386507e-05,
+      "loss": 3.6817,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960574909114018e-05,
+      "loss": 3.6708,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959736314362966e-05,
+      "loss": 3.6499,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958899357492287e-05,
+      "loss": 3.6857,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958060762741235e-05,
+      "loss": 3.6727,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957222167990183e-05,
+      "loss": 3.6763,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956385211119504e-05,
+      "loss": 3.6672,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955546616368452e-05,
+      "loss": 3.6614,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9547080216174e-05,
+      "loss": 3.6608,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953869426866349e-05,
+      "loss": 3.6711,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9530324699956696e-05,
+      "loss": 3.6677,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9521938752446176e-05,
+      "loss": 3.6653,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9513552804935656e-05,
+      "loss": 3.6788,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9505166857425136e-05,
+      "loss": 3.6625,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496780909914616e-05,
+      "loss": 3.6581,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9488411341207825e-05,
+      "loss": 3.6761,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9480025393697305e-05,
+      "loss": 3.6534,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471639446186785e-05,
+      "loss": 3.6585,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9463253498676265e-05,
+      "loss": 3.6735,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454867551165745e-05,
+      "loss": 3.6677,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9446497982458954e-05,
+      "loss": 3.656,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943811203494844e-05,
+      "loss": 3.6591,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942972608743792e-05,
+      "loss": 3.6516,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94213401399274e-05,
+      "loss": 3.6645,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941295419241688e-05,
+      "loss": 3.6735,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940458462371009e-05,
+      "loss": 3.6663,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939619867619957e-05,
+      "loss": 3.6695,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938781272868905e-05,
+      "loss": 3.6761,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937942678117853e-05,
+      "loss": 3.6823,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937104083366801e-05,
+      "loss": 3.6543,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936267126496122e-05,
+      "loss": 3.676,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93542853174507e-05,
+      "loss": 3.6661,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934589936994018e-05,
+      "loss": 3.6493,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933751342242966e-05,
+      "loss": 3.6721,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329143853722875e-05,
+      "loss": 3.6669,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320757906212355e-05,
+      "loss": 3.676,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9312371958701835e-05,
+      "loss": 3.6666,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303986011191315e-05,
+      "loss": 3.6537,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295600063680794e-05,
+      "loss": 3.6546,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287214116170274e-05,
+      "loss": 3.6661,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278844547463484e-05,
+      "loss": 3.6644,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9270458599952964e-05,
+      "loss": 3.6675,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9262072652442443e-05,
+      "loss": 3.6664,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925368670493192e-05,
+      "loss": 3.6638,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9245300757421397e-05,
+      "loss": 3.6652,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236914809910877e-05,
+      "loss": 3.656,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922854524120409e-05,
+      "loss": 3.6591,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922015929369357e-05,
+      "loss": 3.6722,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921177334618305e-05,
+      "loss": 3.6652,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920338739867253e-05,
+      "loss": 3.6729,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919501782996575e-05,
+      "loss": 3.6523,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918663188245522e-05,
+      "loss": 3.6645,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91782459349447e-05,
+      "loss": 3.6681,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916985998743418e-05,
+      "loss": 3.6662,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916149041872739e-05,
+      "loss": 3.6627,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915310447121687e-05,
+      "loss": 3.6615,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914471852370635e-05,
+      "loss": 3.6636,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913633257619583e-05,
+      "loss": 3.6638,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912794662868532e-05,
+      "loss": 3.6522,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91195606811748e-05,
+      "loss": 3.6594,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9111191112468006e-05,
+      "loss": 3.6541,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102805164957486e-05,
+      "loss": 3.6705,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9094419217446966e-05,
+      "loss": 3.6529,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9086033269936446e-05,
+      "loss": 3.6617,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077663701229655e-05,
+      "loss": 3.657,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9069277753719135e-05,
+      "loss": 3.6717,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060891806208615e-05,
+      "loss": 3.6602,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9052505858698095e-05,
+      "loss": 3.6528,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9044136289991304e-05,
+      "loss": 3.6662,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035750342480784e-05,
+      "loss": 3.6721,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027364394970264e-05,
+      "loss": 3.664,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901897844745975e-05,
+      "loss": 3.6612,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901060887875296e-05,
+      "loss": 3.6527,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900222293124244e-05,
+      "loss": 3.6623,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899383698373192e-05,
+      "loss": 3.6622,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89854510362214e-05,
+      "loss": 3.653,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897706508871088e-05,
+      "loss": 3.6546,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896867914120036e-05,
+      "loss": 3.6623,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896030957249357e-05,
+      "loss": 3.6646,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895192362498305e-05,
+      "loss": 3.6629,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894353767747253e-05,
+      "loss": 3.6476,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893515172996201e-05,
+      "loss": 3.663,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892678216125522e-05,
+      "loss": 3.651,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8918396213744704e-05,
+      "loss": 3.6693,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8910010266234184e-05,
+      "loss": 3.6619,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8901624318723664e-05,
+      "loss": 3.6696,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889325475001687e-05,
+      "loss": 3.6451,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888486880250635e-05,
+      "loss": 3.6706,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887648285499583e-05,
+      "loss": 3.6617,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886809690748531e-05,
+      "loss": 3.6665,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885971095997479e-05,
+      "loss": 3.6563,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8851341391268e-05,
+      "loss": 3.6551,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884295544375748e-05,
+      "loss": 3.6642,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883456949624696e-05,
+      "loss": 3.683,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882618354873644e-05,
+      "loss": 3.6552,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881779760122592e-05,
+      "loss": 3.6636,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88094116537154e-05,
+      "loss": 3.6692,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880104208500862e-05,
+      "loss": 3.6537,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87926561374981e-05,
+      "loss": 3.6593,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878427018998758e-05,
+      "loss": 3.6634,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877588424247705e-05,
+      "loss": 3.6676,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876751467377027e-05,
+      "loss": 3.6598,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875912872625975e-05,
+      "loss": 3.6601,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.848252058029175,
+      "eval_runtime": 302.9762,
+      "eval_samples_per_second": 1259.475,
+      "eval_steps_per_second": 39.36,
+      "step": 686880
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.768300209992417e+17,
+  "trial_name": null,
+  "trial_params": null
+}