diff --git "a/checkpoint-534240/trainer_state.json" "b/checkpoint-534240/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-534240/trainer_state.json"
@@ -0,0 +1,6339 @@
+{
+  "best_metric": 3.8663017749786377,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/passive/transformer/1/checkpoints/checkpoint-534240",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 534240,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9323,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.829,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1955,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9876,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8238,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7039,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6145,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5373,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4552,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4069,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3539,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3134,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2658,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2176,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.178,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1336,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.115,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0825,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0539,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0235,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.015,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9821,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9593,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9296,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9278,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8883,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8781,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8482,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8377,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8239,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 4.8065,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740117521192533e-05,
+      "loss": 4.7973,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731731573682013e-05,
+      "loss": 4.784,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7687,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7659,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.746,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7399,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.6986,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7107,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6746,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6858,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6678,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6666,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6418,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.653,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6397,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6337,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6287,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 4.5955,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5991,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6011,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5891,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.5802,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5608,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5591,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5413,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5632,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5192,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5418,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5339,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5138,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5131,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.504,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4957,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4877,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.4991,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4766,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4741,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4664,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4654,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4755,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4707,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4465,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4613,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4609,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4338,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4412,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4307,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4251,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4148,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4198,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4241,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4273,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4024,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3811,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.3965,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.3952,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.3948,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3824,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.3843,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3804,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.377,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3679,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3705,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3589,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3724,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3578,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3395,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3522,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3529,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.356,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3399,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3391,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3242,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3343,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3156,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3236,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.321,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3199,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3109,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3137,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3049,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3124,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.3029,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.2991,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3147,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.3027,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.2995,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2886,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.2941,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2865,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.2855,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.2943,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2783,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2839,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2651,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2693,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2748,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2663,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2585,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2845,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2653,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2586,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2576,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2578,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2585,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2567,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2509,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2584,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2557,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2495,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2407,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.246,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2374,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2389,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.2352,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2332,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.239,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2262,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.2233,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.200882911682129,
+      "eval_runtime": 302.3694,
+      "eval_samples_per_second": 1262.003,
+      "eval_steps_per_second": 39.439,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2084,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.21,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2365,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2209,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.2144,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.2116,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.2083,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.1963,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2087,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.2032,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.206,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.2077,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.1922,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.1882,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.1812,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.1794,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.1899,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.1797,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1838,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.2022,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1809,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1847,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.172,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1883,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1681,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.1691,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.1632,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.1672,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1594,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1576,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.1613,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.1675,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1589,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1619,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.1548,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844918934748812e-05,
+      "loss": 4.1643,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1372,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1529,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1308,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841566193624977e-05,
+      "loss": 4.1516,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1364,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1497,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1317,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382134525011416e-05,
+      "loss": 4.1458,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.144,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.1423,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1457,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834860711377307e-05,
+      "loss": 4.1179,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1238,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.1366,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1342,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831507970253472e-05,
+      "loss": 4.1264,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83066937550242e-05,
+      "loss": 4.1215,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829830780751368e-05,
+      "loss": 4.1168,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828992186000316e-05,
+      "loss": 4.1084,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828155229129637e-05,
+      "loss": 4.1234,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.105,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1149,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.1243,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248024880058025e-05,
+      "loss": 4.1045,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239638932547505e-05,
+      "loss": 4.1043,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231252985036985e-05,
+      "loss": 4.1042,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222867037526465e-05,
+      "loss": 4.0964,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214497468819674e-05,
+      "loss": 4.0984,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206111521309154e-05,
+      "loss": 4.1156,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197725573798634e-05,
+      "loss": 4.0981,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189339626288114e-05,
+      "loss": 4.0959,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818097005758132e-05,
+      "loss": 4.0935,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817258411007081e-05,
+      "loss": 4.0933,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816419816256029e-05,
+      "loss": 4.1088,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815581221504977e-05,
+      "loss": 4.1027,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.0923,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813905669883246e-05,
+      "loss": 4.104,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813067075132194e-05,
+      "loss": 4.1052,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812228480381141e-05,
+      "loss": 4.0876,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811391523510463e-05,
+      "loss": 4.0936,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810552928759411e-05,
+      "loss": 4.0909,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809714334008359e-05,
+      "loss": 4.0888,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808875739257306e-05,
+      "loss": 4.0813,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808037144506255e-05,
+      "loss": 4.0844,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807200187635576e-05,
+      "loss": 4.0962,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063615928845243e-05,
+      "loss": 4.0938,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055229981334717e-05,
+      "loss": 4.0827,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804686041262793e-05,
+      "loss": 4.0602,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803847446511741e-05,
+      "loss": 4.0764,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030088517606886e-05,
+      "loss": 4.0819,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021702570096366e-05,
+      "loss": 4.0806,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801333300138958e-05,
+      "loss": 4.0763,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800494705387906e-05,
+      "loss": 4.0722,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799657748517227e-05,
+      "loss": 4.0791,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798819153766175e-05,
+      "loss": 4.0707,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797980559015123e-05,
+      "loss": 4.0663,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797141964264071e-05,
+      "loss": 4.079,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796303369513019e-05,
+      "loss": 4.0618,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954664126423406e-05,
+      "loss": 4.0825,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946278178912886e-05,
+      "loss": 4.0719,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793789223140236e-05,
+      "loss": 4.0517,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792950628389184e-05,
+      "loss": 4.0694,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792112033638132e-05,
+      "loss": 4.0647,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79127343888708e-05,
+      "loss": 4.075,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790434844136028e-05,
+      "loss": 4.0654,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789596249384976e-05,
+      "loss": 4.0621,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788759292514297e-05,
+      "loss": 4.0499,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787920697763245e-05,
+      "loss": 4.0635,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870821030121935e-05,
+      "loss": 4.0452,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862451461415144e-05,
+      "loss": 4.0532,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854065513904624e-05,
+      "loss": 4.0588,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845679566394104e-05,
+      "loss": 4.0558,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837293618883584e-05,
+      "loss": 4.0446,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828907671373064e-05,
+      "loss": 4.0554,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820521723862544e-05,
+      "loss": 4.049,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812135776352024e-05,
+      "loss": 4.0541,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803749828841504e-05,
+      "loss": 4.0507,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779538026013471e-05,
+      "loss": 4.0429,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778699431262419e-05,
+      "loss": 4.0674,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777860836511367e-05,
+      "loss": 4.0555,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0525,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776185284889637e-05,
+      "loss": 4.0391,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775346690138585e-05,
+      "loss": 4.0466,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774508095387533e-05,
+      "loss": 4.0452,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773669500636481e-05,
+      "loss": 4.0392,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772832543765802e-05,
+      "loss": 4.0547,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77199394901475e-05,
+      "loss": 4.0425,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771155354263698e-05,
+      "loss": 4.0413,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770316759512646e-05,
+      "loss": 4.035,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694798026419666e-05,
+      "loss": 4.0369,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686412078909146e-05,
+      "loss": 4.038,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7678026131398626e-05,
+      "loss": 4.037,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766965656269184e-05,
+      "loss": 4.0266,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766127061518132e-05,
+      "loss": 4.0544,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528846676708e-05,
+      "loss": 4.0437,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764449872016028e-05,
+      "loss": 4.0336,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763611277264976e-05,
+      "loss": 4.0305,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762772682513924e-05,
+      "loss": 4.0299,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761934087762872e-05,
+      "loss": 4.0397,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610954930118195e-05,
+      "loss": 4.0356,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760258536141141e-05,
+      "loss": 4.0309,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759419941390089e-05,
+      "loss": 4.0415,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758581346639037e-05,
+      "loss": 4.0358,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577427518879844e-05,
+      "loss": 4.0378,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756905795017306e-05,
+      "loss": 4.0258,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756067200266254e-05,
+      "loss": 4.0328,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755228605515202e-05,
+      "loss": 4.028,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75439001076415e-05,
+      "loss": 4.0261,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535530538934716e-05,
+      "loss": 4.0267,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527144591424196e-05,
+      "loss": 4.0249,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751875864391367e-05,
+      "loss": 4.0292,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751037269640315e-05,
+      "loss": 4.0176,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502003127696365e-05,
+      "loss": 4.0186,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.032747268676758,
+      "eval_runtime": 302.674,
+      "eval_samples_per_second": 1260.733,
+      "eval_steps_per_second": 39.399,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493617180185845e-05,
+      "loss": 4.0048,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748523123267532e-05,
+      "loss": 4.007,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74768452851648e-05,
+      "loss": 4.0337,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746845933765428e-05,
+      "loss": 4.0255,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460073390143764e-05,
+      "loss": 4.0172,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451687442633244e-05,
+      "loss": 4.0119,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443301495122724e-05,
+      "loss": 4.015,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434915547612204e-05,
+      "loss": 4.0007,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.0185,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.0119,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409757705080644e-05,
+      "loss": 4.0115,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740138813637385e-05,
+      "loss": 4.0153,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 4.002,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 4.0025,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 3.9976,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 3.9954,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 3.9998,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 3.994,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734271926140768e-05,
+      "loss": 4.0013,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0239,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 3.9984,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 4.0038,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 3.9984,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730080590265881e-05,
+      "loss": 4.0074,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729241995514829e-05,
+      "loss": 3.9885,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728403400763777e-05,
+      "loss": 3.9984,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727564806012725e-05,
+      "loss": 3.9898,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267278491420456e-05,
+      "loss": 3.9903,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258892543909936e-05,
+      "loss": 3.9847,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250506596399416e-05,
+      "loss": 3.985,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 3.9899,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723375108018211e-05,
+      "loss": 4.0005,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722536513267159e-05,
+      "loss": 3.9893,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721697918516107e-05,
+      "loss": 3.9968,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 3.9822,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 4.0003,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191821342629505e-05,
+      "loss": 3.97,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 3.9902,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71750658264122e-05,
+      "loss": 3.9657,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716667987890168e-05,
+      "loss": 3.9867,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 3.9753,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149907983880634e-05,
+      "loss": 3.9899,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714153841517385e-05,
+      "loss": 3.9752,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 3.9808,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9836,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9864,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108011003935505e-05,
+      "loss": 3.9837,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709962505642498e-05,
+      "loss": 3.9659,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709123910891446e-05,
+      "loss": 3.9652,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708285316140394e-05,
+      "loss": 3.9821,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074483592697154e-05,
+      "loss": 3.98,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706611402399036e-05,
+      "loss": 3.9719,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705772807647984e-05,
+      "loss": 3.9691,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704934212896932e-05,
+      "loss": 3.9675,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70409561814588e-05,
+      "loss": 3.9536,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703257023394828e-05,
+      "loss": 3.9688,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.9559,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701579833892724e-05,
+      "loss": 3.9672,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700742877022045e-05,
+      "loss": 3.9763,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699904282270993e-05,
+      "loss": 3.9564,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699065687519941e-05,
+      "loss": 3.9572,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698227092768889e-05,
+      "loss": 3.9613,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69739013589821e-05,
+      "loss": 3.9492,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696551541147158e-05,
+      "loss": 3.9492,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695712946396106e-05,
+      "loss": 3.97,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694874351645054e-05,
+      "loss": 3.9539,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694037394774376e-05,
+      "loss": 3.9525,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693198800023324e-05,
+      "loss": 3.9506,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692361843152645e-05,
+      "loss": 3.9528,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915232484015926e-05,
+      "loss": 3.9649,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9618,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9499,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890074641484366e-05,
+      "loss": 3.9648,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881688693973846e-05,
+      "loss": 3.961,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873302746463326e-05,
+      "loss": 3.9504,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864916798952806e-05,
+      "loss": 3.9549,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856547230246015e-05,
+      "loss": 3.9563,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848161282735495e-05,
+      "loss": 3.9511,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683979171402871e-05,
+      "loss": 3.948,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683140576651819e-05,
+      "loss": 3.9472,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682301981900767e-05,
+      "loss": 3.96,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681463387149715e-05,
+      "loss": 3.9609,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680624792398663e-05,
+      "loss": 3.9449,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679786197647611e-05,
+      "loss": 3.9335,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678947602896559e-05,
+      "loss": 3.9396,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678109008145507e-05,
+      "loss": 3.9483,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677272051274828e-05,
+      "loss": 3.9506,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676435094404149e-05,
+      "loss": 3.9422,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675596499653097e-05,
+      "loss": 3.9392,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674757904902045e-05,
+      "loss": 3.9536,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739193101509935e-05,
+      "loss": 3.9364,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730807153999415e-05,
+      "loss": 3.939,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722421206488895e-05,
+      "loss": 3.9469,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714051637782104e-05,
+      "loss": 3.9354,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705665690271584e-05,
+      "loss": 3.9547,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697279742761064e-05,
+      "loss": 3.9413,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688893795250544e-05,
+      "loss": 3.9252,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680507847740024e-05,
+      "loss": 3.942,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672121900229504e-05,
+      "loss": 3.9358,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663735952718984e-05,
+      "loss": 3.952,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665535000520846e-05,
+      "loss": 3.942,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66469968153054e-05,
+      "loss": 3.9341,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663861086779488e-05,
+      "loss": 3.922,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663022492028437e-05,
+      "loss": 3.94,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662183897277385e-05,
+      "loss": 3.9226,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661345302526333e-05,
+      "loss": 3.9328,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660508345655654e-05,
+      "loss": 3.9358,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659669750904602e-05,
+      "loss": 3.9322,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65883115615355e-05,
+      "loss": 3.9218,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657992561402498e-05,
+      "loss": 3.9407,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657153966651446e-05,
+      "loss": 3.9249,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656315371900393e-05,
+      "loss": 3.9292,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554784150297146e-05,
+      "loss": 3.9343,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546398202786626e-05,
+      "loss": 3.9238,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6538012255276106e-05,
+      "loss": 3.9455,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529626307765586e-05,
+      "loss": 3.934,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65212567390588e-05,
+      "loss": 3.9351,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651287079154828e-05,
+      "loss": 3.9219,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650448484403776e-05,
+      "loss": 3.9289,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496098896527235e-05,
+      "loss": 3.9306,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487712949016715e-05,
+      "loss": 3.9174,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479327001506195e-05,
+      "loss": 3.9337,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470941053995675e-05,
+      "loss": 3.9266,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462555106485155e-05,
+      "loss": 3.9249,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454185537778364e-05,
+      "loss": 3.9247,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445799590267844e-05,
+      "loss": 3.9238,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437413642757324e-05,
+      "loss": 3.9162,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429027695246804e-05,
+      "loss": 3.9262,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642065812654002e-05,
+      "loss": 3.9156,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64122721790295e-05,
+      "loss": 3.9364,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640388623151898e-05,
+      "loss": 3.9335,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639550028400846e-05,
+      "loss": 3.9185,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638711433649794e-05,
+      "loss": 3.9168,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637872838898742e-05,
+      "loss": 3.918,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63703424414769e-05,
+      "loss": 3.9255,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636195649396638e-05,
+      "loss": 3.9251,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635360330406332e-05,
+      "loss": 3.9197,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63452173565528e-05,
+      "loss": 3.9264,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633683140904228e-05,
+      "loss": 3.9247,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632844546153176e-05,
+      "loss": 3.9284,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320059514021244e-05,
+      "loss": 3.9127,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311689945314454e-05,
+      "loss": 3.9265,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303303997803934e-05,
+      "loss": 3.9211,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294918050293413e-05,
+      "loss": 3.9114,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286532102782893e-05,
+      "loss": 3.9212,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278146155272373e-05,
+      "loss": 3.9196,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626976020776185e-05,
+      "loss": 3.9171,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626137426025133e-05,
+      "loss": 3.9079,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625300469154454e-05,
+      "loss": 3.9093,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.957947015762329,
+      "eval_runtime": 310.3378,
+      "eval_samples_per_second": 1229.599,
+      "eval_steps_per_second": 38.426,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624461874403402e-05,
+      "loss": 3.901,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.62362327965235e-05,
+      "loss": 3.8992,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622784684901298e-05,
+      "loss": 3.9255,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621946090150246e-05,
+      "loss": 3.9197,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621107495399194e-05,
+      "loss": 3.9194,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620268900648142e-05,
+      "loss": 3.8999,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61943030589709e-05,
+      "loss": 3.9147,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618591711146038e-05,
+      "loss": 3.8927,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617753116394986e-05,
+      "loss": 3.9172,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616914521643934e-05,
+      "loss": 3.9098,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616077564773255e-05,
+      "loss": 3.9031,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615238970022203e-05,
+      "loss": 3.9158,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614402013151524e-05,
+      "loss": 3.8985,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613563418400472e-05,
+      "loss": 3.9035,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61272482364942e-05,
+      "loss": 3.892,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611886228898368e-05,
+      "loss": 3.8996,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611047634147316e-05,
+      "loss": 3.8963,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610209039396265e-05,
+      "loss": 3.8959,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.609370444645213e-05,
+      "loss": 3.8991,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085318498941607e-05,
+      "loss": 3.9207,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6076932551431087e-05,
+      "loss": 3.8986,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6068546603920566e-05,
+      "loss": 3.9052,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6060160656410046e-05,
+      "loss": 3.9,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6051791087703256e-05,
+      "loss": 3.9076,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6043405140192735e-05,
+      "loss": 3.8913,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6035019192682215e-05,
+      "loss": 3.8992,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6026633245171695e-05,
+      "loss": 3.8914,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6018263676464904e-05,
+      "loss": 3.8905,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6009877728954384e-05,
+      "loss": 3.8858,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6001491781443864e-05,
+      "loss": 3.8938,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5993105833933344e-05,
+      "loss": 3.8913,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598473626522656e-05,
+      "loss": 3.9036,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597636669651977e-05,
+      "loss": 3.8932,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596798074900925e-05,
+      "loss": 3.8972,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595959480149873e-05,
+      "loss": 3.8882,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595120885398821e-05,
+      "loss": 3.9021,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594282290647769e-05,
+      "loss": 3.8821,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593443695896717e-05,
+      "loss": 3.8923,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.592606739026038e-05,
+      "loss": 3.8715,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591768144274986e-05,
+      "loss": 3.8905,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590929549523934e-05,
+      "loss": 3.8848,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590090954772882e-05,
+      "loss": 3.8975,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58925236002183e-05,
+      "loss": 3.8854,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5884137652707785e-05,
+      "loss": 3.8876,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587575170519726e-05,
+      "loss": 3.8905,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5867382136490474e-05,
+      "loss": 3.8962,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5858996188979954e-05,
+      "loss": 3.886,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585061024146943e-05,
+      "loss": 3.8771,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584222429395891e-05,
+      "loss": 3.8686,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583383834644839e-05,
+      "loss": 3.8916,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58254687777416e-05,
+      "loss": 3.8904,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5817082830231076e-05,
+      "loss": 3.882,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5808696882720556e-05,
+      "loss": 3.8755,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5800310935210036e-05,
+      "loss": 3.8802,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579192498769952e-05,
+      "loss": 3.8612,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783539040189e-05,
+      "loss": 3.8842,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577516947148221e-05,
+      "loss": 3.8649,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576678352397169e-05,
+      "loss": 3.8776,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575839757646117e-05,
+      "loss": 3.8841,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575001162895065e-05,
+      "loss": 3.8748,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574162568144013e-05,
+      "loss": 3.8652,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.573323973392961e-05,
+      "loss": 3.874,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.572485378641909e-05,
+      "loss": 3.8614,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.571646783890857e-05,
+      "loss": 3.8658,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.570809827020178e-05,
+      "loss": 3.8799,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569971232269126e-05,
+      "loss": 3.869,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569132637518074e-05,
+      "loss": 3.8647,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5682956806473956e-05,
+      "loss": 3.8605,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674570858963436e-05,
+      "loss": 3.8672,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5666184911452916e-05,
+      "loss": 3.8721,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5657798963942396e-05,
+      "loss": 3.8751,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5649413016431876e-05,
+      "loss": 3.8682,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5641027068921356e-05,
+      "loss": 3.8781,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5632641121410836e-05,
+      "loss": 3.8752,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5624271552704045e-05,
+      "loss": 3.8649,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5615885605193525e-05,
+      "loss": 3.8673,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5607499657683005e-05,
+      "loss": 3.8748,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5599113710172485e-05,
+      "loss": 3.87,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5590744141465694e-05,
+      "loss": 3.8634,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5582358193955174e-05,
+      "loss": 3.8641,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5573972246444654e-05,
+      "loss": 3.8711,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556558629893414e-05,
+      "loss": 3.8742,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557200351423614e-05,
+      "loss": 3.8654,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548814403913094e-05,
+      "loss": 3.8515,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540428456402574e-05,
+      "loss": 3.8513,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5532042508892054e-05,
+      "loss": 3.8625,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552367294018526e-05,
+      "loss": 3.8699,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551530337147848e-05,
+      "loss": 3.8647,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550691742396796e-05,
+      "loss": 3.8515,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549853147645744e-05,
+      "loss": 3.8701,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549014552894691e-05,
+      "loss": 3.8593,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548175958143639e-05,
+      "loss": 3.8571,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547339001272961e-05,
+      "loss": 3.8603,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546500406521909e-05,
+      "loss": 3.8629,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545661811770857e-05,
+      "loss": 3.8741,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544823217019805e-05,
+      "loss": 3.8549,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543984622268753e-05,
+      "loss": 3.8441,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5431476653980736e-05,
+      "loss": 3.8629,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5423090706470216e-05,
+      "loss": 3.8548,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5414704758959696e-05,
+      "loss": 3.8692,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5406318811449176e-05,
+      "loss": 3.8642,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5397949242742385e-05,
+      "loss": 3.8521,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389563295231865e-05,
+      "loss": 3.8445,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5381177347721345e-05,
+      "loss": 3.8613,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.537279140021083e-05,
+      "loss": 3.8438,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536440545270031e-05,
+      "loss": 3.8539,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535603588399352e-05,
+      "loss": 3.8522,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5347649936483e-05,
+      "loss": 3.8579,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533926398897248e-05,
+      "loss": 3.8423,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533087804146196e-05,
+      "loss": 3.8602,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532249209395144e-05,
+      "loss": 3.8458,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531410614644092e-05,
+      "loss": 3.851,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.530573657773413e-05,
+      "loss": 3.8592,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529735063022361e-05,
+      "loss": 3.8461,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528896468271309e-05,
+      "loss": 3.8673,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528057873520257e-05,
+      "loss": 3.858,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5272209166495786e-05,
+      "loss": 3.8567,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5263823218985266e-05,
+      "loss": 3.8453,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255437271474746e-05,
+      "loss": 3.8511,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5247051323964226e-05,
+      "loss": 3.8491,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238665376453706e-05,
+      "loss": 3.8445,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5230295807746915e-05,
+      "loss": 3.8537,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5221909860236395e-05,
+      "loss": 3.8503,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5213523912725875e-05,
+      "loss": 3.8497,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5205137965215355e-05,
+      "loss": 3.8462,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5196752017704834e-05,
+      "loss": 3.8505,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5188366070194314e-05,
+      "loss": 3.8419,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5179996501487524e-05,
+      "loss": 3.8526,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5171610553977003e-05,
+      "loss": 3.8398,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5163224606466483e-05,
+      "loss": 3.8594,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515483865895597e-05,
+      "loss": 3.8528,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514645271144545e-05,
+      "loss": 3.8506,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513808314273866e-05,
+      "loss": 3.8393,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512969719522814e-05,
+      "loss": 3.8447,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512131124771762e-05,
+      "loss": 3.8507,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51129253002071e-05,
+      "loss": 3.8519,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510453935269657e-05,
+      "loss": 3.8416,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509616978398979e-05,
+      "loss": 3.8505,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508778383647927e-05,
+      "loss": 3.8478,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507939788896874e-05,
+      "loss": 3.8572,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507101194145822e-05,
+      "loss": 3.8426,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506264237275144e-05,
+      "loss": 3.8533,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5054256425240924e-05,
+      "loss": 3.8509,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50458704777304e-05,
+      "loss": 3.8401,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503748453021988e-05,
+      "loss": 3.8448,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502909858270936e-05,
+      "loss": 3.8525,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502072901400257e-05,
+      "loss": 3.8438,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5012343066492046e-05,
+      "loss": 3.8307,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5003957118981526e-05,
+      "loss": 3.8464,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.917541742324829,
+      "eval_runtime": 304.6162,
+      "eval_samples_per_second": 1252.694,
+      "eval_steps_per_second": 39.148,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4995571171471006e-05,
+      "loss": 3.833,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4987185223960486e-05,
+      "loss": 3.8275,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978799276449966e-05,
+      "loss": 3.8502,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4970429707743175e-05,
+      "loss": 3.8502,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.496204376023266e-05,
+      "loss": 3.8486,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.495365781272214e-05,
+      "loss": 3.8358,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494527186521162e-05,
+      "loss": 3.8419,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49368859177011e-05,
+      "loss": 3.8207,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492849997019058e-05,
+      "loss": 3.8464,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492011402268006e-05,
+      "loss": 3.8413,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491174445397327e-05,
+      "loss": 3.833,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490335850646275e-05,
+      "loss": 3.8456,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489498893775596e-05,
+      "loss": 3.8287,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.488660299024544e-05,
+      "loss": 3.8342,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487821704273492e-05,
+      "loss": 3.8255,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.48698310952244e-05,
+      "loss": 3.8262,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486144514771388e-05,
+      "loss": 3.8294,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485305920020336e-05,
+      "loss": 3.8263,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.484467325269284e-05,
+      "loss": 3.8303,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4836287305182326e-05,
+      "loss": 3.8474,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4827901357671806e-05,
+      "loss": 3.8327,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4819515410161286e-05,
+      "loss": 3.8386,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4811145841454495e-05,
+      "loss": 3.8306,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4802759893943975e-05,
+      "loss": 3.8368,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4794373946433455e-05,
+      "loss": 3.8243,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4785987998922935e-05,
+      "loss": 3.831,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.477760205141241e-05,
+      "loss": 3.8286,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4769232482705624e-05,
+      "loss": 3.8227,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4760846535195104e-05,
+      "loss": 3.8174,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475246058768458e-05,
+      "loss": 3.8269,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4744074640174064e-05,
+      "loss": 3.8273,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473570507146728e-05,
+      "loss": 3.8372,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472731912395676e-05,
+      "loss": 3.827,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471893317644623e-05,
+      "loss": 3.8326,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471054722893571e-05,
+      "loss": 3.8199,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470217766022893e-05,
+      "loss": 3.8361,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469379171271841e-05,
+      "loss": 3.8169,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468540576520788e-05,
+      "loss": 3.8244,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467701981769736e-05,
+      "loss": 3.8077,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466865024899058e-05,
+      "loss": 3.8231,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466026430148005e-05,
+      "loss": 3.8201,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.465187835396953e-05,
+      "loss": 3.8304,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.464349240645902e-05,
+      "loss": 3.8182,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.46351064589485e-05,
+      "loss": 3.8249,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.462672051143798e-05,
+      "loss": 3.8244,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618350942731186e-05,
+      "loss": 3.8287,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609964995220666e-05,
+      "loss": 3.825,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4601579047710146e-05,
+      "loss": 3.8127,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4593193100199626e-05,
+      "loss": 3.8019,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4584807152689106e-05,
+      "loss": 3.8294,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576437583982315e-05,
+      "loss": 3.8268,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4568051636471795e-05,
+      "loss": 3.8171,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4559665688961275e-05,
+      "loss": 3.8152,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4551279741450755e-05,
+      "loss": 3.8167,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4542893793940235e-05,
+      "loss": 3.7976,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453452422523345e-05,
+      "loss": 3.8138,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452613827772293e-05,
+      "loss": 3.8058,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451775233021241e-05,
+      "loss": 3.8124,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450936638270189e-05,
+      "loss": 3.8233,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450098043519137e-05,
+      "loss": 3.8124,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.449259448768085e-05,
+      "loss": 3.8006,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.448420854017033e-05,
+      "loss": 3.8159,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.447583897146354e-05,
+      "loss": 3.7971,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446745302395302e-05,
+      "loss": 3.8057,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44590670764425e-05,
+      "loss": 3.816,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445068112893198e-05,
+      "loss": 3.8094,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.444231156022519e-05,
+      "loss": 3.8021,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.443392561271467e-05,
+      "loss": 3.8025,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4425539665204156e-05,
+      "loss": 3.8018,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4417153717693636e-05,
+      "loss": 3.8113,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4408784148986845e-05,
+      "loss": 3.8115,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4400398201476325e-05,
+      "loss": 3.8075,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4392012253965805e-05,
+      "loss": 3.8163,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4383626306455285e-05,
+      "loss": 3.8156,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4375240358944764e-05,
+      "loss": 3.8045,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4366870790237974e-05,
+      "loss": 3.8004,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4358484842727454e-05,
+      "loss": 3.8193,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4350098895216933e-05,
+      "loss": 3.8091,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4341712947706413e-05,
+      "loss": 3.7989,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433332700019589e-05,
+      "loss": 3.8044,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43249574314891e-05,
+      "loss": 3.8113,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431657148397859e-05,
+      "loss": 3.818,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.430818553646807e-05,
+      "loss": 3.8004,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429979958895754e-05,
+      "loss": 3.7904,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429141364144702e-05,
+      "loss": 3.798,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428304407274024e-05,
+      "loss": 3.7974,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.427465812522971e-05,
+      "loss": 3.8079,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.426627217771919e-05,
+      "loss": 3.8049,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.425788623020867e-05,
+      "loss": 3.7922,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424951666150189e-05,
+      "loss": 3.8093,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424113071399136e-05,
+      "loss": 3.7992,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423274476648084e-05,
+      "loss": 3.8011,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422435881897033e-05,
+      "loss": 3.7979,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.421598925026354e-05,
+      "loss": 3.8016,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4207603302753016e-05,
+      "loss": 3.8152,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4199217355242496e-05,
+      "loss": 3.8024,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4190831407731976e-05,
+      "loss": 3.7825,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4182461839025185e-05,
+      "loss": 3.803,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4174075891514665e-05,
+      "loss": 3.7975,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4165689944004145e-05,
+      "loss": 3.8054,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4157303996493625e-05,
+      "loss": 3.8087,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4148934427786834e-05,
+      "loss": 3.7959,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4140548480276314e-05,
+      "loss": 3.7873,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4132162532765794e-05,
+      "loss": 3.8008,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.412377658525528e-05,
+      "loss": 3.7826,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411539063774476e-05,
+      "loss": 3.7983,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.410702106903797e-05,
+      "loss": 3.7948,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409863512152745e-05,
+      "loss": 3.8019,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409024917401693e-05,
+      "loss": 3.7846,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408186322650641e-05,
+      "loss": 3.8022,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407347727899589e-05,
+      "loss": 3.7879,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40651077102891e-05,
+      "loss": 3.7923,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405672176277858e-05,
+      "loss": 3.8028,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404833581526806e-05,
+      "loss": 3.7939,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403994986775754e-05,
+      "loss": 3.8087,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403158029905075e-05,
+      "loss": 3.8007,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4023194351540234e-05,
+      "loss": 3.7956,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4014808404029714e-05,
+      "loss": 3.7969,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4006422456519194e-05,
+      "loss": 3.7899,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39980528878124e-05,
+      "loss": 3.7976,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398966694030188e-05,
+      "loss": 3.788,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398128099279136e-05,
+      "loss": 3.801,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397289504528084e-05,
+      "loss": 3.7901,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396452547657405e-05,
+      "loss": 3.7902,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.395613952906353e-05,
+      "loss": 3.7899,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394775358155301e-05,
+      "loss": 3.7974,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393936763404249e-05,
+      "loss": 3.7865,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39309980653357e-05,
+      "loss": 3.7963,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.392261211782519e-05,
+      "loss": 3.7867,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.391422617031467e-05,
+      "loss": 3.8017,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390584022280415e-05,
+      "loss": 3.7964,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389745427529363e-05,
+      "loss": 3.796,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388908470658684e-05,
+      "loss": 3.7862,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388069875907632e-05,
+      "loss": 3.7872,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38723128115658e-05,
+      "loss": 3.801,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.386392686405528e-05,
+      "loss": 3.793,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3855557295348486e-05,
+      "loss": 3.7857,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3847171347837966e-05,
+      "loss": 3.7995,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3838785400327446e-05,
+      "loss": 3.7934,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3830399452816926e-05,
+      "loss": 3.8055,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.382202988411014e-05,
+      "loss": 3.7895,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381364393659962e-05,
+      "loss": 3.7964,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38052579890891e-05,
+      "loss": 3.7949,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379687204157858e-05,
+      "loss": 3.7854,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378848609406806e-05,
+      "loss": 3.7925,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378011652536127e-05,
+      "loss": 3.7935,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377173057785075e-05,
+      "loss": 3.7934,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.376334463034023e-05,
+      "loss": 3.7779,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.375495868282971e-05,
+      "loss": 3.7922,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.892491102218628,
+      "eval_runtime": 304.2151,
+      "eval_samples_per_second": 1254.346,
+      "eval_steps_per_second": 39.199,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374657273531919e-05,
+      "loss": 3.7787,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373818678780867e-05,
+      "loss": 3.773,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3729800840298144e-05,
+      "loss": 3.7954,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721414892787624e-05,
+      "loss": 3.7981,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371302894527711e-05,
+      "loss": 3.7928,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370464299776659e-05,
+      "loss": 3.7835,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369625705025607e-05,
+      "loss": 3.7863,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368787110274555e-05,
+      "loss": 3.77,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367950153403876e-05,
+      "loss": 3.7909,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367111558652824e-05,
+      "loss": 3.7904,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366272963901772e-05,
+      "loss": 3.7787,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36543436915072e-05,
+      "loss": 3.7903,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3645990501604144e-05,
+      "loss": 3.7779,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363760455409362e-05,
+      "loss": 3.7824,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36292186065831e-05,
+      "loss": 3.7731,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362083265907258e-05,
+      "loss": 3.7749,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3612446711562064e-05,
+      "loss": 3.7761,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3604060764051544e-05,
+      "loss": 3.7767,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595674816541024e-05,
+      "loss": 3.7792,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3587288869030504e-05,
+      "loss": 3.7929,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357891930032371e-05,
+      "loss": 3.7863,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357053335281319e-05,
+      "loss": 3.7862,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356214740530267e-05,
+      "loss": 3.7789,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355376145779215e-05,
+      "loss": 3.7856,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354537551028163e-05,
+      "loss": 3.7709,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353700594157484e-05,
+      "loss": 3.7846,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352861999406432e-05,
+      "loss": 3.7748,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35202340465538e-05,
+      "loss": 3.7704,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351184809904328e-05,
+      "loss": 3.7699,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35034785303365e-05,
+      "loss": 3.7726,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349509258282598e-05,
+      "loss": 3.779,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348670663531546e-05,
+      "loss": 3.7822,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347832068780494e-05,
+      "loss": 3.7811,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3469951119098146e-05,
+      "loss": 3.7822,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461565171587626e-05,
+      "loss": 3.7681,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3453179224077106e-05,
+      "loss": 3.7833,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444793276566586e-05,
+      "loss": 3.7701,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436407329056066e-05,
+      "loss": 3.7682,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3428037760349275e-05,
+      "loss": 3.7643,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419651812838755e-05,
+      "loss": 3.7716,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411265865328235e-05,
+      "loss": 3.7738,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402879917817715e-05,
+      "loss": 3.7796,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339451034911093e-05,
+      "loss": 3.7699,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338612440160041e-05,
+      "loss": 3.7708,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337773845408989e-05,
+      "loss": 3.7782,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336935250657937e-05,
+      "loss": 3.7768,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336098293787258e-05,
+      "loss": 3.7768,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335259699036206e-05,
+      "loss": 3.7637,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334421104285154e-05,
+      "loss": 3.7483,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333582509534102e-05,
+      "loss": 3.7816,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332745552663423e-05,
+      "loss": 3.7776,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331906957912371e-05,
+      "loss": 3.77,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331068363161319e-05,
+      "loss": 3.7685,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330229768410267e-05,
+      "loss": 3.7628,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3293928115395885e-05,
+      "loss": 3.7518,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285542167885365e-05,
+      "loss": 3.7633,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3277156220374845e-05,
+      "loss": 3.7558,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268770272864325e-05,
+      "loss": 3.7639,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3260384325353805e-05,
+      "loss": 3.7764,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3252014756647014e-05,
+      "loss": 3.7616,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243628809136494e-05,
+      "loss": 3.7526,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3235242861625974e-05,
+      "loss": 3.77,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322685691411545e-05,
+      "loss": 3.7496,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321848734540866e-05,
+      "loss": 3.7547,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321010139789814e-05,
+      "loss": 3.7706,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320171545038762e-05,
+      "loss": 3.7632,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31933295028771e-05,
+      "loss": 3.7545,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318495993417032e-05,
+      "loss": 3.7535,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31765739866598e-05,
+      "loss": 3.754,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316818803914928e-05,
+      "loss": 3.7613,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315980209163875e-05,
+      "loss": 3.7679,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315143252293197e-05,
+      "loss": 3.7563,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314304657542145e-05,
+      "loss": 3.7683,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313466062791092e-05,
+      "loss": 3.7673,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3126291059204136e-05,
+      "loss": 3.7663,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117905111693616e-05,
+      "loss": 3.7492,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109519164183096e-05,
+      "loss": 3.77,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101133216672576e-05,
+      "loss": 3.7647,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309276364796579e-05,
+      "loss": 3.7491,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308437770045527e-05,
+      "loss": 3.7628,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307599175294475e-05,
+      "loss": 3.7602,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067605805434225e-05,
+      "loss": 3.7705,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3059219857923705e-05,
+      "loss": 3.7548,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305085028921692e-05,
+      "loss": 3.7415,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042464341706394e-05,
+      "loss": 3.7525,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034078394195874e-05,
+      "loss": 3.7534,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025692446685354e-05,
+      "loss": 3.7597,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301732287797857e-05,
+      "loss": 3.757,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300893693046805e-05,
+      "loss": 3.7489,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300055098295753e-05,
+      "loss": 3.7575,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299216503544701e-05,
+      "loss": 3.7562,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983795466740226e-05,
+      "loss": 3.7492,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29754095192297e-05,
+      "loss": 3.7562,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296702357171918e-05,
+      "loss": 3.7603,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295863762420866e-05,
+      "loss": 3.7688,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295026805550187e-05,
+      "loss": 3.7538,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294188210799135e-05,
+      "loss": 3.7397,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293349616048083e-05,
+      "loss": 3.756,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292511021297031e-05,
+      "loss": 3.751,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291674064426352e-05,
+      "loss": 3.7618,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2908354696753004e-05,
+      "loss": 3.7645,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2899968749242484e-05,
+      "loss": 3.7479,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2891582801731964e-05,
+      "loss": 3.7443,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288321323302517e-05,
+      "loss": 3.7572,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287482728551465e-05,
+      "loss": 3.7369,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286644133800413e-05,
+      "loss": 3.756,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285805539049361e-05,
+      "loss": 3.7432,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284968582178682e-05,
+      "loss": 3.759,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28412998742763e-05,
+      "loss": 3.7405,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283291392676578e-05,
+      "loss": 3.7547,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282452797925526e-05,
+      "loss": 3.7485,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281615841054847e-05,
+      "loss": 3.7443,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280777246303796e-05,
+      "loss": 3.7587,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279938651552744e-05,
+      "loss": 3.7471,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279100056801692e-05,
+      "loss": 3.7644,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2782630999310126e-05,
+      "loss": 3.7603,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2774245051799606e-05,
+      "loss": 3.7499,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2765859104289086e-05,
+      "loss": 3.7555,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2757473156778566e-05,
+      "loss": 3.7458,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2749103588071775e-05,
+      "loss": 3.7495,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2740717640561255e-05,
+      "loss": 3.7476,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732331693050735e-05,
+      "loss": 3.754,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2723945745540215e-05,
+      "loss": 3.7474,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715576176833424e-05,
+      "loss": 3.746,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2707190229322904e-05,
+      "loss": 3.7486,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269880428181239e-05,
+      "loss": 3.7509,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269041833430187e-05,
+      "loss": 3.7421,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268204876559508e-05,
+      "loss": 3.7503,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267366281808456e-05,
+      "loss": 3.7401,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266527687057404e-05,
+      "loss": 3.7645,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265689092306352e-05,
+      "loss": 3.7452,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264852135435673e-05,
+      "loss": 3.7566,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264013540684621e-05,
+      "loss": 3.742,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263174945933569e-05,
+      "loss": 3.7465,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262336351182517e-05,
+      "loss": 3.7558,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261499394311838e-05,
+      "loss": 3.7489,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260660799560786e-05,
+      "loss": 3.7439,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2598222048097345e-05,
+      "loss": 3.7576,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589836100586825e-05,
+      "loss": 3.7497,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2581466531880034e-05,
+      "loss": 3.7607,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2573080584369514e-05,
+      "loss": 3.7438,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564694636858994e-05,
+      "loss": 3.7528,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2556308689348474e-05,
+      "loss": 3.7508,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254793912064168e-05,
+      "loss": 3.7434,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253955317313116e-05,
+      "loss": 3.7516,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253116722562064e-05,
+      "loss": 3.7512,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252278127811012e-05,
+      "loss": 3.7548,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251441170940333e-05,
+      "loss": 3.7268,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250602576189281e-05,
+      "loss": 3.7538,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.87673020362854,
+      "eval_runtime": 304.8371,
+      "eval_samples_per_second": 1251.787,
+      "eval_steps_per_second": 39.119,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24976398143823e-05,
+      "loss": 3.74,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248925386687178e-05,
+      "loss": 3.7313,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248088429816499e-05,
+      "loss": 3.7502,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247249835065447e-05,
+      "loss": 3.7589,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246411240314395e-05,
+      "loss": 3.7499,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245572645563343e-05,
+      "loss": 3.739,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2447356886926636e-05,
+      "loss": 3.7448,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2438970939416116e-05,
+      "loss": 3.7287,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2430584991905596e-05,
+      "loss": 3.7462,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2422199044395076e-05,
+      "loss": 3.7457,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2413829475688285e-05,
+      "loss": 3.7449,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2405443528177765e-05,
+      "loss": 3.7456,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239707395947098e-05,
+      "loss": 3.7403,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238868801196046e-05,
+      "loss": 3.7378,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238030206444994e-05,
+      "loss": 3.7329,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237191611693942e-05,
+      "loss": 3.7279,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23635301694289e-05,
+      "loss": 3.7354,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235514422191838e-05,
+      "loss": 3.7348,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234675827440786e-05,
+      "loss": 3.7369,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233837232689734e-05,
+      "loss": 3.7497,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233000275819055e-05,
+      "loss": 3.7475,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232161681068003e-05,
+      "loss": 3.7421,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231323086316951e-05,
+      "loss": 3.739,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230484491565899e-05,
+      "loss": 3.7433,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2296475346952206e-05,
+      "loss": 3.7283,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2288089399441686e-05,
+      "loss": 3.7453,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2279703451931166e-05,
+      "loss": 3.7339,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2271317504420645e-05,
+      "loss": 3.7244,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2262947935713855e-05,
+      "loss": 3.7328,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2254561988203335e-05,
+      "loss": 3.7321,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2246176040692814e-05,
+      "loss": 3.7351,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2237806471986024e-05,
+      "loss": 3.7414,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2229420524475504e-05,
+      "loss": 3.7391,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2221034576964983e-05,
+      "loss": 3.7404,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2212648629454463e-05,
+      "loss": 3.7275,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2204262681943943e-05,
+      "loss": 3.7476,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219587673443342e-05,
+      "loss": 3.7294,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21874907869229e-05,
+      "loss": 3.7227,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217910483941238e-05,
+      "loss": 3.7294,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21707352707056e-05,
+      "loss": 3.7257,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216234932319507e-05,
+      "loss": 3.7326,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215396337568455e-05,
+      "loss": 3.7425,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214557742817403e-05,
+      "loss": 3.7288,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213720785946725e-05,
+      "loss": 3.7307,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212882191195672e-05,
+      "loss": 3.7383,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21204359644462e-05,
+      "loss": 3.7402,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211206639573942e-05,
+      "loss": 3.7385,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21036804482289e-05,
+      "loss": 3.725,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209529450071838e-05,
+      "loss": 3.701,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208690855320786e-05,
+      "loss": 3.7439,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207852260569734e-05,
+      "loss": 3.7373,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2070153036990546e-05,
+      "loss": 3.7332,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061767089480026e-05,
+      "loss": 3.7261,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2053381141969506e-05,
+      "loss": 3.724,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204501157326272e-05,
+      "loss": 3.713,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036625625752195e-05,
+      "loss": 3.7226,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2028239678241675e-05,
+      "loss": 3.7196,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019853730731155e-05,
+      "loss": 3.7235,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011484162024364e-05,
+      "loss": 3.7391,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200309821451385e-05,
+      "loss": 3.7183,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199471226700333e-05,
+      "loss": 3.7136,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198632631949281e-05,
+      "loss": 3.731,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197795675078602e-05,
+      "loss": 3.7151,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19695708032755e-05,
+      "loss": 3.7155,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196118485576498e-05,
+      "loss": 3.7263,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195279890825446e-05,
+      "loss": 3.7232,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194441296074394e-05,
+      "loss": 3.7135,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193604339203715e-05,
+      "loss": 3.7147,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192765744452663e-05,
+      "loss": 3.7149,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191927149701611e-05,
+      "loss": 3.7262,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191090192830932e-05,
+      "loss": 3.7257,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1902515980798804e-05,
+      "loss": 3.7215,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1894130033288284e-05,
+      "loss": 3.7248,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1885744085777764e-05,
+      "loss": 3.7291,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1877374517070973e-05,
+      "loss": 3.7285,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186898856956045e-05,
+      "loss": 3.7139,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186060262204993e-05,
+      "loss": 3.7283,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.185221667453941e-05,
+      "loss": 3.7293,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184384710583262e-05,
+      "loss": 3.708,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18354611583221e-05,
+      "loss": 3.7263,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182707521081158e-05,
+      "loss": 3.7232,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181868926330106e-05,
+      "loss": 3.7283,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181031969459427e-05,
+      "loss": 3.7187,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180193374708375e-05,
+      "loss": 3.704,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179354779957324e-05,
+      "loss": 3.7122,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.178516185206272e-05,
+      "loss": 3.7155,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177679228335593e-05,
+      "loss": 3.7243,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176840633584541e-05,
+      "loss": 3.7142,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176002038833489e-05,
+      "loss": 3.7107,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175163444082437e-05,
+      "loss": 3.7196,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1743264872117576e-05,
+      "loss": 3.7232,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1734878924607056e-05,
+      "loss": 3.7088,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1726492977096536e-05,
+      "loss": 3.7154,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1718107029586016e-05,
+      "loss": 3.7246,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1709737460879225e-05,
+      "loss": 3.7308,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1701351513368705e-05,
+      "loss": 3.718,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169296556585819e-05,
+      "loss": 3.7043,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168457961834767e-05,
+      "loss": 3.7183,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167621004964088e-05,
+      "loss": 3.7107,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166782410213036e-05,
+      "loss": 3.7245,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165943815461984e-05,
+      "loss": 3.7279,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165105220710932e-05,
+      "loss": 3.7129,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164268263840253e-05,
+      "loss": 3.7066,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163429669089201e-05,
+      "loss": 3.7157,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162591074338149e-05,
+      "loss": 3.7008,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161752479587097e-05,
+      "loss": 3.7196,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160915522716418e-05,
+      "loss": 3.7003,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160076927965366e-05,
+      "loss": 3.7217,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1592383332143145e-05,
+      "loss": 3.7051,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1583997384632625e-05,
+      "loss": 3.7151,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1575627815925834e-05,
+      "loss": 3.711,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1567241868415314e-05,
+      "loss": 3.7119,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1558855920904794e-05,
+      "loss": 3.7201,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1550469973394274e-05,
+      "loss": 3.71,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154210040468748e-05,
+      "loss": 3.7281,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153371445717696e-05,
+      "loss": 3.7222,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152532850966644e-05,
+      "loss": 3.715,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151694256215592e-05,
+      "loss": 3.7182,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150857299344913e-05,
+      "loss": 3.7081,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150018704593861e-05,
+      "loss": 3.7136,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14918010984281e-05,
+      "loss": 3.7121,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.148341515091758e-05,
+      "loss": 3.7101,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147504558221079e-05,
+      "loss": 3.7125,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146665963470027e-05,
+      "loss": 3.7111,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145827368718975e-05,
+      "loss": 3.7123,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144988773967923e-05,
+      "loss": 3.7132,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144151817097244e-05,
+      "loss": 3.7084,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143313222346192e-05,
+      "loss": 3.714,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14247462759514e-05,
+      "loss": 3.7011,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141636032844088e-05,
+      "loss": 3.7257,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1407990759734086e-05,
+      "loss": 3.7142,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1399604812223566e-05,
+      "loss": 3.721,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139121886471305e-05,
+      "loss": 3.701,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138283291720253e-05,
+      "loss": 3.7145,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.137446334849574e-05,
+      "loss": 3.7178,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.136607740098522e-05,
+      "loss": 3.7154,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13576914534747e-05,
+      "loss": 3.7042,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134930550596418e-05,
+      "loss": 3.7222,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134093593725739e-05,
+      "loss": 3.712,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133254998974687e-05,
+      "loss": 3.7274,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132416404223635e-05,
+      "loss": 3.7059,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131577809472583e-05,
+      "loss": 3.721,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130740852601904e-05,
+      "loss": 3.7135,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129902257850852e-05,
+      "loss": 3.7092,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1290636630998006e-05,
+      "loss": 3.7134,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1282250683487486e-05,
+      "loss": 3.7161,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1273881114780695e-05,
+      "loss": 3.7175,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1265495167270175e-05,
+      "loss": 3.6947,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1257109219759655e-05,
+      "loss": 3.7207,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8663017749786377,
+      "eval_runtime": 302.6553,
+      "eval_samples_per_second": 1260.811,
+      "eval_steps_per_second": 39.401,
+      "step": 534240
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.7187113081337856e+17,
+  "trial_name": null,
+  "trial_params": null
+}