diff --git "a/checkpoint-457920/trainer_state.json" "b/checkpoint-457920/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457920/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 3.891617774963379,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/transformer/4/checkpoints/checkpoint-457920",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 457920,
+  "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": 10.9521,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8262,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1939,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9899,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8315,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.734,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6212,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5598,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4784,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4148,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3678,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3295,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2839,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2291,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1923,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1557,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1189,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0961,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0691,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0371,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0251,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.9935,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9745,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9472,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9372,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.9066,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8872,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8746,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8532,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8294,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.824,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740117521192533e-05,
+      "loss": 4.8135,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731731573682013e-05,
+      "loss": 4.7994,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7811,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.777,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.754,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7444,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7168,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7181,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.7048,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.7026,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6893,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6854,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6668,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.6605,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6566,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6392,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6453,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6222,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95892359812181e-05,
+      "loss": 4.6061,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.6182,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5989,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.5903,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5709,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5828,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5577,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5893,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5521,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5646,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5548,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.53,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 4.5234,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.528,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.5125,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.506,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 4.5236,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.507,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4937,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4786,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4766,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.487,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.489,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4771,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4754,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4726,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4661,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.454,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4516,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4477,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4351,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4384,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4397,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4394,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.4221,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.4017,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.4219,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.4163,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.4146,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.4065,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.4078,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3966,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3884,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3916,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3893,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3966,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3941,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3859,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3678,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3704,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3724,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3715,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3604,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3594,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3481,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3645,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3332,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3462,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3458,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3349,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3354,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3367,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3286,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3344,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3296,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3226,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3259,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.33,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.3205,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.3099,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.3134,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.314,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.3069,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.2962,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.3048,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.3082,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2993,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2944,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2866,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2897,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2842,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2976,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2894,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2701,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2808,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2843,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.2759,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2816,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2675,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2764,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2789,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2674,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8818040007521146e-05,
+      "loss": 4.2642,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2654,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2532,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2532,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.266,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2455,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2504,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2546,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.2507,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.215391635894775,
+      "eval_runtime": 310.9131,
+      "eval_samples_per_second": 1227.324,
+      "eval_steps_per_second": 38.355,
+      "step": 76320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2277,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2349,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.253,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2354,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2476,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2215,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.2376,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.2233,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.2222,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.2242,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.2245,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.2279,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.2123,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.2083,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.2065,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.2035,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.202,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.2046,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.2018,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.858328261363778e-05,
+      "loss": 4.2275,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.857489666612726e-05,
+      "loss": 4.2,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.2051,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.1988,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.2047,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1937,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.1918,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.1903,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.1845,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.1698,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1863,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1834,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.1882,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1787,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1829,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.1718,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1724,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844078702117387e-05,
+      "loss": 4.1591,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843240107366335e-05,
+      "loss": 4.1669,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.842401512615282e-05,
+      "loss": 4.1633,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84156291786423e-05,
+      "loss": 4.1638,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.1644,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.1699,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.1592,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.1607,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.1635,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.1495,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.1611,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.1478,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.1404,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.1568,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8323416513634046e-05,
+      "loss": 4.1439,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.1408,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.1337,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.1415,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.1263,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.1564,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8273133586178384e-05,
+      "loss": 4.1318,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.1453,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.1426,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.824797574364683e-05,
+      "loss": 4.1245,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.1192,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.1306,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.1206,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.821446471121221e-05,
+      "loss": 4.1166,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.820607876370169e-05,
+      "loss": 4.1337,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.819769281619117e-05,
+      "loss": 4.1317,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818930686868065e-05,
+      "loss": 4.1154,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818093729997386e-05,
+      "loss": 4.1098,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817255135246334e-05,
+      "loss": 4.1121,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816416540495282e-05,
+      "loss": 4.1219,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8155779457442305e-05,
+      "loss": 4.1236,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.1217,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8139023941224994e-05,
+      "loss": 4.1226,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8130637993714474e-05,
+      "loss": 4.1226,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8122252046203954e-05,
+      "loss": 4.1236,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8113866098693434e-05,
+      "loss": 4.1056,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.810549652998664e-05,
+      "loss": 4.1138,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.809711058247612e-05,
+      "loss": 4.1126,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.80887246349656e-05,
+      "loss": 4.0995,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808033868745508e-05,
+      "loss": 4.1112,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.807196911874829e-05,
+      "loss": 4.1137,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.806358317123777e-05,
+      "loss": 4.1145,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.805519722372726e-05,
+      "loss": 4.0972,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804681127621674e-05,
+      "loss": 4.0862,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803844170750995e-05,
+      "loss": 4.103,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803005575999943e-05,
+      "loss": 4.1014,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802166981248891e-05,
+      "loss": 4.1048,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801328386497839e-05,
+      "loss": 4.0976,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8004914296271596e-05,
+      "loss": 4.0983,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7996528348761076e-05,
+      "loss": 4.0983,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7988142401250556e-05,
+      "loss": 4.0844,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7979756453740036e-05,
+      "loss": 4.0928,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7971370506229516e-05,
+      "loss": 4.0927,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7962984558718996e-05,
+      "loss": 4.1017,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7954598611208476e-05,
+      "loss": 4.1035,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7946212663697956e-05,
+      "loss": 4.0972,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793784309499117e-05,
+      "loss": 4.0823,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792945714748065e-05,
+      "loss": 4.0875,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792107119997013e-05,
+      "loss": 4.0865,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7912685252459605e-05,
+      "loss": 4.0953,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.790431568375282e-05,
+      "loss": 4.0823,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.78959297362423e-05,
+      "loss": 4.0804,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7887543788731774e-05,
+      "loss": 4.0732,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7879157841221254e-05,
+      "loss": 4.0938,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787078827251447e-05,
+      "loss": 4.0612,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.786240232500395e-05,
+      "loss": 4.0773,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7854032756297166e-05,
+      "loss": 4.0838,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845646808786646e-05,
+      "loss": 4.0714,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837260861276126e-05,
+      "loss": 4.0724,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7828874913765606e-05,
+      "loss": 4.0809,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782048896625508e-05,
+      "loss": 4.071,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.781210301874456e-05,
+      "loss": 4.0731,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780371707123404e-05,
+      "loss": 4.0769,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.779534750252725e-05,
+      "loss": 4.0702,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778696155501673e-05,
+      "loss": 4.0772,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777857560750621e-05,
+      "loss": 4.0821,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7770206038799423e-05,
+      "loss": 4.0731,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7761820091288903e-05,
+      "loss": 4.0653,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7753434143778383e-05,
+      "loss": 4.0699,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774504819626786e-05,
+      "loss": 4.0673,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773666224875734e-05,
+      "loss": 4.0636,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.772827630124682e-05,
+      "loss": 4.0588,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77198903537363e-05,
+      "loss": 4.0619,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771150440622578e-05,
+      "loss": 4.0748,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.770313483751899e-05,
+      "loss": 4.0638,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769474889000847e-05,
+      "loss": 4.0637,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.768636294249795e-05,
+      "loss": 4.0532,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.767799337379116e-05,
+      "loss": 4.0561,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766960742628064e-05,
+      "loss": 4.0539,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766122147877012e-05,
+      "loss": 4.0708,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.76528355312596e-05,
+      "loss": 4.0662,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764444958374909e-05,
+      "loss": 4.044,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.76360800150423e-05,
+      "loss": 4.0523,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762769406753178e-05,
+      "loss": 4.0622,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761930812002126e-05,
+      "loss": 4.0547,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761092217251074e-05,
+      "loss": 4.0627,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7602552603803946e-05,
+      "loss": 4.0484,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7594166656293426e-05,
+      "loss": 4.0548,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7585780708782906e-05,
+      "loss": 4.0605,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7577394761272386e-05,
+      "loss": 4.05,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7569025192565595e-05,
+      "loss": 4.0513,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7560639245055075e-05,
+      "loss": 4.0532,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7552253297544555e-05,
+      "loss": 4.0434,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7543867350034035e-05,
+      "loss": 4.042,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.753549778132725e-05,
+      "loss": 4.0553,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.752711183381673e-05,
+      "loss": 4.0371,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751872588630621e-05,
+      "loss": 4.0451,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751035631759942e-05,
+      "loss": 4.0469,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75019703700889e-05,
+      "loss": 4.0435,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.046647071838379,
+      "eval_runtime": 311.1083,
+      "eval_samples_per_second": 1226.554,
+      "eval_steps_per_second": 38.331,
+      "step": 152640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.749358442257838e-05,
+      "loss": 4.0327,
+      "step": 153088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.748519847506786e-05,
+      "loss": 4.0312,
+      "step": 153600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.747682890636107e-05,
+      "loss": 4.0519,
+      "step": 154112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.746844295885055e-05,
+      "loss": 4.033,
+      "step": 154624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.746005701134003e-05,
+      "loss": 4.0513,
+      "step": 155136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745167106382951e-05,
+      "loss": 4.0273,
+      "step": 155648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7443301495122724e-05,
+      "loss": 4.0415,
+      "step": 156160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7434915547612204e-05,
+      "loss": 4.0236,
+      "step": 156672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.0375,
+      "step": 157184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.0319,
+      "step": 157696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.740977408388437e-05,
+      "loss": 4.0307,
+      "step": 158208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.740138813637385e-05,
+      "loss": 4.0398,
+      "step": 158720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 4.0228,
+      "step": 159232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 4.0213,
+      "step": 159744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 4.0197,
+      "step": 160256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 4.0152,
+      "step": 160768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 4.02,
+      "step": 161280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 4.0234,
+      "step": 161792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.734271926140768e-05,
+      "loss": 4.0184,
+      "step": 162304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0446,
+      "step": 162816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 4.0219,
+      "step": 163328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 4.027,
+      "step": 163840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 4.0213,
+      "step": 164352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730080590265881e-05,
+      "loss": 4.0271,
+      "step": 164864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7292436333952016e-05,
+      "loss": 4.0184,
+      "step": 165376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7284050386441496e-05,
+      "loss": 4.0156,
+      "step": 165888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7275664438930976e-05,
+      "loss": 4.0186,
+      "step": 166400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7267278491420456e-05,
+      "loss": 4.0088,
+      "step": 166912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7258892543909936e-05,
+      "loss": 3.9972,
+      "step": 167424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7250506596399416e-05,
+      "loss": 4.0124,
+      "step": 167936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 4.0198,
+      "step": 168448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.723375108018211e-05,
+      "loss": 4.0134,
+      "step": 168960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.722536513267159e-05,
+      "loss": 4.0135,
+      "step": 169472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.721697918516107e-05,
+      "loss": 4.0162,
+      "step": 169984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 4.0041,
+      "step": 170496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 4.0056,
+      "step": 171008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7191821342629505e-05,
+      "loss": 3.9981,
+      "step": 171520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 4.0008,
+      "step": 172032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.71750658264122e-05,
+      "loss": 3.9971,
+      "step": 172544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.716667987890168e-05,
+      "loss": 4.0007,
+      "step": 173056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 4.0031,
+      "step": 173568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7149907983880634e-05,
+      "loss": 4.0095,
+      "step": 174080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714153841517385e-05,
+      "loss": 3.9988,
+      "step": 174592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 4.0011,
+      "step": 175104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 4.0069,
+      "step": 175616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9893,
+      "step": 176128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7108011003935505e-05,
+      "loss": 4.0018,
+      "step": 176640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7099641435228714e-05,
+      "loss": 3.9958,
+      "step": 177152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7091255487718194e-05,
+      "loss": 3.9801,
+      "step": 177664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7082869540207674e-05,
+      "loss": 3.9993,
+      "step": 178176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7074483592697154e-05,
+      "loss": 3.9926,
+      "step": 178688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.706609764518663e-05,
+      "loss": 3.9863,
+      "step": 179200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.705771169767611e-05,
+      "loss": 3.9854,
+      "step": 179712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704932575016559e-05,
+      "loss": 3.9889,
+      "step": 180224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.70409561814588e-05,
+      "loss": 3.9718,
+      "step": 180736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.703257023394828e-05,
+      "loss": 3.9995,
+      "step": 181248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.984,
+      "step": 181760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.701579833892724e-05,
+      "loss": 3.9949,
+      "step": 182272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.700742877022045e-05,
+      "loss": 3.9958,
+      "step": 182784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.699904282270993e-05,
+      "loss": 3.9784,
+      "step": 183296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.699065687519941e-05,
+      "loss": 3.9725,
+      "step": 183808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.698227092768889e-05,
+      "loss": 3.9873,
+      "step": 184320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.69739013589821e-05,
+      "loss": 3.9731,
+      "step": 184832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.696551541147158e-05,
+      "loss": 3.9704,
+      "step": 185344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.695712946396106e-05,
+      "loss": 3.9896,
+      "step": 185856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694874351645054e-05,
+      "loss": 3.989,
+      "step": 186368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694037394774376e-05,
+      "loss": 3.9721,
+      "step": 186880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.693200437903697e-05,
+      "loss": 3.9668,
+      "step": 187392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.692361843152645e-05,
+      "loss": 3.9691,
+      "step": 187904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6915232484015926e-05,
+      "loss": 3.9776,
+      "step": 188416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9826,
+      "step": 188928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9799,
+      "step": 189440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6890074641484366e-05,
+      "loss": 3.9864,
+      "step": 189952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6881688693973846e-05,
+      "loss": 3.9788,
+      "step": 190464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6873302746463326e-05,
+      "loss": 3.9868,
+      "step": 190976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6864933177756535e-05,
+      "loss": 3.9687,
+      "step": 191488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6856547230246015e-05,
+      "loss": 3.976,
+      "step": 192000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6848161282735495e-05,
+      "loss": 3.9782,
+      "step": 192512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.683977533522498e-05,
+      "loss": 3.9617,
+      "step": 193024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.683140576651819e-05,
+      "loss": 3.9745,
+      "step": 193536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.682301981900767e-05,
+      "loss": 3.9822,
+      "step": 194048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.681463387149715e-05,
+      "loss": 3.9793,
+      "step": 194560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.680624792398663e-05,
+      "loss": 3.9633,
+      "step": 195072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.679787835527984e-05,
+      "loss": 3.9512,
+      "step": 195584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678950878657305e-05,
+      "loss": 3.9693,
+      "step": 196096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678112283906253e-05,
+      "loss": 3.9614,
+      "step": 196608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.677273689155201e-05,
+      "loss": 3.9752,
+      "step": 197120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.676435094404149e-05,
+      "loss": 3.9674,
+      "step": 197632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.675596499653097e-05,
+      "loss": 3.9663,
+      "step": 198144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.674757904902045e-05,
+      "loss": 3.9661,
+      "step": 198656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6739209480313664e-05,
+      "loss": 3.9531,
+      "step": 199168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6730823532803144e-05,
+      "loss": 3.9639,
+      "step": 199680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6722437585292624e-05,
+      "loss": 3.9619,
+      "step": 200192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6714051637782104e-05,
+      "loss": 3.9754,
+      "step": 200704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6705665690271584e-05,
+      "loss": 3.9767,
+      "step": 201216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.669729612156479e-05,
+      "loss": 3.9678,
+      "step": 201728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668891017405427e-05,
+      "loss": 3.9526,
+      "step": 202240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668052422654375e-05,
+      "loss": 3.9628,
+      "step": 202752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.667213827903323e-05,
+      "loss": 3.9596,
+      "step": 203264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.666376871032644e-05,
+      "loss": 3.9689,
+      "step": 203776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.665538276281592e-05,
+      "loss": 3.9552,
+      "step": 204288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.66469968153054e-05,
+      "loss": 3.9609,
+      "step": 204800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.663861086779488e-05,
+      "loss": 3.9477,
+      "step": 205312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.663022492028437e-05,
+      "loss": 3.9694,
+      "step": 205824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.662183897277385e-05,
+      "loss": 3.9395,
+      "step": 206336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.661345302526333e-05,
+      "loss": 3.9569,
+      "step": 206848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.660508345655654e-05,
+      "loss": 3.9525,
+      "step": 207360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.659669750904602e-05,
+      "loss": 3.9535,
+      "step": 207872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.65883115615355e-05,
+      "loss": 3.9485,
+      "step": 208384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.657992561402498e-05,
+      "loss": 3.9583,
+      "step": 208896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6571556045318187e-05,
+      "loss": 3.957,
+      "step": 209408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6563170097807667e-05,
+      "loss": 3.9495,
+      "step": 209920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6554784150297146e-05,
+      "loss": 3.9584,
+      "step": 210432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6546398202786626e-05,
+      "loss": 3.945,
+      "step": 210944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6538028634079836e-05,
+      "loss": 3.9632,
+      "step": 211456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652964268656932e-05,
+      "loss": 3.9656,
+      "step": 211968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652127311786253e-05,
+      "loss": 3.9494,
+      "step": 212480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.651288717035201e-05,
+      "loss": 3.9473,
+      "step": 212992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.650450122284149e-05,
+      "loss": 3.9496,
+      "step": 213504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.649611527533097e-05,
+      "loss": 3.9493,
+      "step": 214016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.648772932782045e-05,
+      "loss": 3.9442,
+      "step": 214528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647934338030993e-05,
+      "loss": 3.9441,
+      "step": 215040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6470957432799404e-05,
+      "loss": 3.9407,
+      "step": 215552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6462571485288884e-05,
+      "loss": 3.9588,
+      "step": 216064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6454185537778364e-05,
+      "loss": 3.9502,
+      "step": 216576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6445799590267844e-05,
+      "loss": 3.9466,
+      "step": 217088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.643743002156106e-05,
+      "loss": 3.9386,
+      "step": 217600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.642904407405054e-05,
+      "loss": 3.9433,
+      "step": 218112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.642065812654002e-05,
+      "loss": 3.9371,
+      "step": 218624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.64122721790295e-05,
+      "loss": 3.9583,
+      "step": 219136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.640390261032271e-05,
+      "loss": 3.9533,
+      "step": 219648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.639551666281219e-05,
+      "loss": 3.935,
+      "step": 220160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.638713071530167e-05,
+      "loss": 3.9359,
+      "step": 220672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637874476779115e-05,
+      "loss": 3.95,
+      "step": 221184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637035882028063e-05,
+      "loss": 3.9411,
+      "step": 221696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.636198925157384e-05,
+      "loss": 3.9529,
+      "step": 222208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.635360330406332e-05,
+      "loss": 3.9349,
+      "step": 222720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.63452173565528e-05,
+      "loss": 3.9482,
+      "step": 223232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.633683140904228e-05,
+      "loss": 3.9436,
+      "step": 223744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.632844546153176e-05,
+      "loss": 3.9447,
+      "step": 224256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6320075892824974e-05,
+      "loss": 3.9419,
+      "step": 224768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6311689945314454e-05,
+      "loss": 3.9416,
+      "step": 225280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6303303997803934e-05,
+      "loss": 3.9338,
+      "step": 225792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6294918050293413e-05,
+      "loss": 3.9367,
+      "step": 226304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6286532102782893e-05,
+      "loss": 3.9452,
+      "step": 226816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.62781625340761e-05,
+      "loss": 3.9343,
+      "step": 227328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626977658656558e-05,
+      "loss": 3.9327,
+      "step": 227840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626139063905506e-05,
+      "loss": 3.9366,
+      "step": 228352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.625300469154454e-05,
+      "loss": 3.936,
+      "step": 228864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.972496509552002,
+      "eval_runtime": 310.4496,
+      "eval_samples_per_second": 1229.156,
+      "eval_steps_per_second": 38.412,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624461874403402e-05,
+      "loss": 3.9217,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.62362327965235e-05,
+      "loss": 3.9254,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622784684901298e-05,
+      "loss": 3.9456,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621946090150246e-05,
+      "loss": 3.9274,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621107495399194e-05,
+      "loss": 3.9492,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620268900648142e-05,
+      "loss": 3.9189,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61943030589709e-05,
+      "loss": 3.9386,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618591711146038e-05,
+      "loss": 3.923,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617754754275359e-05,
+      "loss": 3.9291,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616916159524307e-05,
+      "loss": 3.9294,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616077564773255e-05,
+      "loss": 3.9265,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615238970022203e-05,
+      "loss": 3.9348,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6144036510318976e-05,
+      "loss": 3.9207,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6135650562808456e-05,
+      "loss": 3.919,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6127264615297936e-05,
+      "loss": 3.923,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118878667787416e-05,
+      "loss": 3.9097,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110492720276896e-05,
+      "loss": 3.9217,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6102106772766376e-05,
+      "loss": 3.9227,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093720825255856e-05,
+      "loss": 3.9173,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085334877745336e-05,
+      "loss": 3.9409,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6076965309038545e-05,
+      "loss": 3.925,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6068579361528025e-05,
+      "loss": 3.9313,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6060193414017505e-05,
+      "loss": 3.9201,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6051807466506985e-05,
+      "loss": 3.9252,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6043437897800194e-05,
+      "loss": 3.9186,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6035051950289674e-05,
+      "loss": 3.9183,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6026666002779154e-05,
+      "loss": 3.9209,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6018280055268634e-05,
+      "loss": 3.9092,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6009894107758114e-05,
+      "loss": 3.9056,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600152453905133e-05,
+      "loss": 3.9104,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599313859154081e-05,
+      "loss": 3.9226,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598475264403029e-05,
+      "loss": 3.9192,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597636669651977e-05,
+      "loss": 3.9207,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596799712781298e-05,
+      "loss": 3.9225,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595961118030246e-05,
+      "loss": 3.9037,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595122523279194e-05,
+      "loss": 3.9113,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594283928528142e-05,
+      "loss": 3.9082,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593446971657463e-05,
+      "loss": 3.9012,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.592608376906411e-05,
+      "loss": 3.9038,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591769782155359e-05,
+      "loss": 3.9035,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590931187404307e-05,
+      "loss": 3.9113,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590094230533628e-05,
+      "loss": 3.9138,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589255635782576e-05,
+      "loss": 3.9101,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588417041031524e-05,
+      "loss": 3.9032,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587578446280472e-05,
+      "loss": 3.9155,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58673985152942e-05,
+      "loss": 3.8973,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585902894658741e-05,
+      "loss": 3.9085,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585064299907689e-05,
+      "loss": 3.9034,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584225705156637e-05,
+      "loss": 3.8887,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583388748285958e-05,
+      "loss": 3.9101,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582550153534906e-05,
+      "loss": 3.9018,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581711558783854e-05,
+      "loss": 3.9,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580872964032802e-05,
+      "loss": 3.8919,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580034369281751e-05,
+      "loss": 3.8971,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579197412411072e-05,
+      "loss": 3.8835,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57835881766002e-05,
+      "loss": 3.9064,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577520222908968e-05,
+      "loss": 3.8969,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576681628157916e-05,
+      "loss": 3.9018,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5758430334068637e-05,
+      "loss": 3.9077,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5750060765361846e-05,
+      "loss": 3.8916,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741674817851326e-05,
+      "loss": 3.8833,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5733288870340806e-05,
+      "loss": 3.9025,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724902922830286e-05,
+      "loss": 3.8844,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716533354123495e-05,
+      "loss": 3.8822,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5708147406612975e-05,
+      "loss": 3.9011,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569976145910246e-05,
+      "loss": 3.8981,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569137551159194e-05,
+      "loss": 3.8866,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568300594288515e-05,
+      "loss": 3.8839,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.567461999537463e-05,
+      "loss": 3.8768,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.566623404786411e-05,
+      "loss": 3.8869,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565784810035359e-05,
+      "loss": 3.8984,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56494785316468e-05,
+      "loss": 3.8971,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564109258413628e-05,
+      "loss": 3.8999,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563270663662576e-05,
+      "loss": 3.8909,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562432068911524e-05,
+      "loss": 3.907,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561595112040845e-05,
+      "loss": 3.8818,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560756517289793e-05,
+      "loss": 3.8937,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5599179225387415e-05,
+      "loss": 3.8888,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5590793277876895e-05,
+      "loss": 3.8754,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5582407330366375e-05,
+      "loss": 3.8969,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5574037761659584e-05,
+      "loss": 3.8972,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5565651814149064e-05,
+      "loss": 3.8899,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557265866638544e-05,
+      "loss": 3.8799,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548879919128024e-05,
+      "loss": 3.872,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554051035042123e-05,
+      "loss": 3.8837,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553212440291071e-05,
+      "loss": 3.8843,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552373845540019e-05,
+      "loss": 3.8908,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5515352507889666e-05,
+      "loss": 3.8801,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550698293918288e-05,
+      "loss": 3.8875,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549859699167237e-05,
+      "loss": 3.8804,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549021104416185e-05,
+      "loss": 3.8762,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548182509665132e-05,
+      "loss": 3.879,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547345552794454e-05,
+      "loss": 3.8761,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546506958043402e-05,
+      "loss": 3.8947,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54566836329235e-05,
+      "loss": 3.8919,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544829768541297e-05,
+      "loss": 3.8919,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543992811670619e-05,
+      "loss": 3.8697,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5431542169195667e-05,
+      "loss": 3.8834,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542315622168514e-05,
+      "loss": 3.8794,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541477027417462e-05,
+      "loss": 3.885,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5406400705467836e-05,
+      "loss": 3.878,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539801475795732e-05,
+      "loss": 3.8783,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389628810446796e-05,
+      "loss": 3.8685,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5381242862936275e-05,
+      "loss": 3.8893,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.537287329422949e-05,
+      "loss": 3.8616,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536448734671897e-05,
+      "loss": 3.8763,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5356101399208444e-05,
+      "loss": 3.8748,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5347715451697924e-05,
+      "loss": 3.8738,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533934588299114e-05,
+      "loss": 3.8693,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5330959935480613e-05,
+      "loss": 3.8843,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5322573987970093e-05,
+      "loss": 3.8775,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5314188040459573e-05,
+      "loss": 3.87,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.530580209294906e-05,
+      "loss": 3.8833,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529743252424227e-05,
+      "loss": 3.8692,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528904657673175e-05,
+      "loss": 3.8814,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528066062922123e-05,
+      "loss": 3.8877,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5272291060514445e-05,
+      "loss": 3.8764,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526390511300392e-05,
+      "loss": 3.8712,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52555191654934e-05,
+      "loss": 3.8692,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524713321798288e-05,
+      "loss": 3.8695,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523874727047236e-05,
+      "loss": 3.8679,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523036132296184e-05,
+      "loss": 3.8735,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522197537545132e-05,
+      "loss": 3.8586,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52135894279408e-05,
+      "loss": 3.8814,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520521985923401e-05,
+      "loss": 3.877,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5196833911723494e-05,
+      "loss": 3.8732,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5188447964212974e-05,
+      "loss": 3.8604,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5180062016702454e-05,
+      "loss": 3.8665,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517169244799566e-05,
+      "loss": 3.8641,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516330650048514e-05,
+      "loss": 3.878,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515492055297462e-05,
+      "loss": 3.8808,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51465346054641e-05,
+      "loss": 3.8664,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513816503675731e-05,
+      "loss": 3.8539,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512977908924679e-05,
+      "loss": 3.8741,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512139314173627e-05,
+      "loss": 3.8653,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.511300719422575e-05,
+      "loss": 3.8803,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510463762551896e-05,
+      "loss": 3.8593,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509625167800845e-05,
+      "loss": 3.8721,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508786573049793e-05,
+      "loss": 3.8709,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507947978298741e-05,
+      "loss": 3.8751,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5071110214280616e-05,
+      "loss": 3.8701,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5062724266770096e-05,
+      "loss": 3.867,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5054338319259576e-05,
+      "loss": 3.8605,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5045952371749056e-05,
+      "loss": 3.8637,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5037582803042265e-05,
+      "loss": 3.8686,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5029196855531745e-05,
+      "loss": 3.8658,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5020810908021225e-05,
+      "loss": 3.8593,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5012424960510705e-05,
+      "loss": 3.8691,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5004055391803914e-05,
+      "loss": 3.8613,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9313888549804688,
+      "eval_runtime": 306.8209,
+      "eval_samples_per_second": 1243.693,
+      "eval_steps_per_second": 38.866,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49956694442934e-05,
+      "loss": 3.8555,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.498728349678288e-05,
+      "loss": 3.8555,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497889754927236e-05,
+      "loss": 3.8705,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497052798056557e-05,
+      "loss": 3.8593,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.496214203305505e-05,
+      "loss": 3.873,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.495375608554453e-05,
+      "loss": 3.8557,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494537013803401e-05,
+      "loss": 3.8616,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.493700056932722e-05,
+      "loss": 3.8545,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49286146218167e-05,
+      "loss": 3.8587,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492022867430618e-05,
+      "loss": 3.8582,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491184272679566e-05,
+      "loss": 3.8573,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490347315808887e-05,
+      "loss": 3.8651,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4895103589382084e-05,
+      "loss": 3.8536,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4886717641871564e-05,
+      "loss": 3.849,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4878331694361044e-05,
+      "loss": 3.8506,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4869945746850524e-05,
+      "loss": 3.8408,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4861559799340004e-05,
+      "loss": 3.8506,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4853173851829484e-05,
+      "loss": 3.8518,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4844787904318964e-05,
+      "loss": 3.8495,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4836401956808444e-05,
+      "loss": 3.8714,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.482803238810165e-05,
+      "loss": 3.8558,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481964644059113e-05,
+      "loss": 3.8614,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481126049308061e-05,
+      "loss": 3.8525,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.480287454557009e-05,
+      "loss": 3.8561,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.479450497686331e-05,
+      "loss": 3.8466,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.478611902935279e-05,
+      "loss": 3.8556,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.477773308184227e-05,
+      "loss": 3.8514,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476934713433175e-05,
+      "loss": 3.8448,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476097756562496e-05,
+      "loss": 3.8405,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475259161811444e-05,
+      "loss": 3.8438,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474420567060392e-05,
+      "loss": 3.8511,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4735836101897126e-05,
+      "loss": 3.8545,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4727450154386606e-05,
+      "loss": 3.8505,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4719064206876086e-05,
+      "loss": 3.8533,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4710678259365566e-05,
+      "loss": 3.8387,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4702292311855046e-05,
+      "loss": 3.8461,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4693906364344526e-05,
+      "loss": 3.8431,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4685520416834006e-05,
+      "loss": 3.8338,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4677134469323486e-05,
+      "loss": 3.8408,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4668764900616695e-05,
+      "loss": 3.8351,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4660378953106175e-05,
+      "loss": 3.847,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4651993005595655e-05,
+      "loss": 3.8489,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.464362343688887e-05,
+      "loss": 3.8442,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4635237489378344e-05,
+      "loss": 3.8358,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4626851541867824e-05,
+      "loss": 3.8504,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618465594357304e-05,
+      "loss": 3.8324,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4610079646846784e-05,
+      "loss": 3.8458,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460171007814e-05,
+      "loss": 3.838,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.459332413062948e-05,
+      "loss": 3.8229,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.458493818311896e-05,
+      "loss": 3.8477,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.457656861441217e-05,
+      "loss": 3.8371,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.456818266690165e-05,
+      "loss": 3.8354,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455979671939113e-05,
+      "loss": 3.831,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455141077188061e-05,
+      "loss": 3.8288,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.454304120317382e-05,
+      "loss": 3.8257,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45346552556633e-05,
+      "loss": 3.8375,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452626930815278e-05,
+      "loss": 3.8373,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451788336064226e-05,
+      "loss": 3.8365,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450951379193547e-05,
+      "loss": 3.8422,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4501127844424954e-05,
+      "loss": 3.8306,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4492741896914433e-05,
+      "loss": 3.8184,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4484355949403913e-05,
+      "loss": 3.8401,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.447598638069712e-05,
+      "loss": 3.8232,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44676004331866e-05,
+      "loss": 3.8165,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445921448567608e-05,
+      "loss": 3.8426,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445082853816556e-05,
+      "loss": 3.8343,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.444245896945877e-05,
+      "loss": 3.8263,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.443407302194825e-05,
+      "loss": 3.8228,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.442568707443773e-05,
+      "loss": 3.8134,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.441730112692721e-05,
+      "loss": 3.825,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440893155822042e-05,
+      "loss": 3.8403,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44005456107099e-05,
+      "loss": 3.8326,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.439215966319939e-05,
+      "loss": 3.8358,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.438377371568887e-05,
+      "loss": 3.8293,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4375404146982076e-05,
+      "loss": 3.8459,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4367018199471556e-05,
+      "loss": 3.8215,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4358632251961036e-05,
+      "loss": 3.8344,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4350246304450516e-05,
+      "loss": 3.8293,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4341876735743725e-05,
+      "loss": 3.8098,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4333490788233205e-05,
+      "loss": 3.8406,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4325104840722685e-05,
+      "loss": 3.8351,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4316718893212165e-05,
+      "loss": 3.8314,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4308349324505374e-05,
+      "loss": 3.8213,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4299963376994854e-05,
+      "loss": 3.8099,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429157742948434e-05,
+      "loss": 3.8259,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428320786077755e-05,
+      "loss": 3.8224,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.427482191326703e-05,
+      "loss": 3.8309,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.426643596575651e-05,
+      "loss": 3.8223,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.425805001824599e-05,
+      "loss": 3.8257,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424966407073547e-05,
+      "loss": 3.8216,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424127812322495e-05,
+      "loss": 3.8206,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423289217571443e-05,
+      "loss": 3.8193,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422452260700764e-05,
+      "loss": 3.8172,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.421613665949712e-05,
+      "loss": 3.8323,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.42077507119866e-05,
+      "loss": 3.8329,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419936476447608e-05,
+      "loss": 3.8374,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4190995195769294e-05,
+      "loss": 3.8094,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4182609248258774e-05,
+      "loss": 3.825,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4174223300748254e-05,
+      "loss": 3.8219,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4165837353237734e-05,
+      "loss": 3.8247,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4157467784530943e-05,
+      "loss": 3.8219,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4149081837020423e-05,
+      "loss": 3.8166,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.41406958895099e-05,
+      "loss": 3.8154,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.413230994199938e-05,
+      "loss": 3.8294,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.412394037329259e-05,
+      "loss": 3.8035,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411555442578207e-05,
+      "loss": 3.8174,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.410716847827155e-05,
+      "loss": 3.8115,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409878253076103e-05,
+      "loss": 3.8206,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409041296205425e-05,
+      "loss": 3.8125,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408202701454373e-05,
+      "loss": 3.824,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407364106703321e-05,
+      "loss": 3.8217,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406525511952269e-05,
+      "loss": 3.8131,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40568855508159e-05,
+      "loss": 3.8252,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404849960330538e-05,
+      "loss": 3.8131,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404011365579486e-05,
+      "loss": 3.8264,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403172770828434e-05,
+      "loss": 3.8281,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4023358139577546e-05,
+      "loss": 3.817,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4014972192067026e-05,
+      "loss": 3.8145,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4006586244556506e-05,
+      "loss": 3.8168,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3998200297045986e-05,
+      "loss": 3.8113,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39898307283392e-05,
+      "loss": 3.8137,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398144478082868e-05,
+      "loss": 3.817,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397305883331816e-05,
+      "loss": 3.804,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396467288580764e-05,
+      "loss": 3.8237,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.395630331710085e-05,
+      "loss": 3.8193,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394791736959033e-05,
+      "loss": 3.8199,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393953142207981e-05,
+      "loss": 3.8051,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393114547456929e-05,
+      "loss": 3.8111,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39227759058625e-05,
+      "loss": 3.8103,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.391438995835198e-05,
+      "loss": 3.8197,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390600401084146e-05,
+      "loss": 3.8226,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389761806333094e-05,
+      "loss": 3.8139,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3889248494624155e-05,
+      "loss": 3.7994,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3880862547113635e-05,
+      "loss": 3.8197,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3872476599603115e-05,
+      "loss": 3.8103,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3864090652092595e-05,
+      "loss": 3.8227,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3855721083385804e-05,
+      "loss": 3.805,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3847335135875284e-05,
+      "loss": 3.8204,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3838949188364764e-05,
+      "loss": 3.8159,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3830563240854244e-05,
+      "loss": 3.8217,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3822193672147453e-05,
+      "loss": 3.8135,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381380772463693e-05,
+      "loss": 3.8125,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.380543815593014e-05,
+      "loss": 3.8064,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379705220841962e-05,
+      "loss": 3.811,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378866626090911e-05,
+      "loss": 3.815,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378028031339859e-05,
+      "loss": 3.8113,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377189436588807e-05,
+      "loss": 3.8064,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.376350841837754e-05,
+      "loss": 3.81,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.375512247086702e-05,
+      "loss": 3.8119,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.906362533569336,
+      "eval_runtime": 307.2937,
+      "eval_samples_per_second": 1241.78,
+      "eval_steps_per_second": 38.807,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37467365233565e-05,
+      "loss": 3.7979,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373835057584598e-05,
+      "loss": 3.7982,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372996462833546e-05,
+      "loss": 3.8167,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372157868082494e-05,
+      "loss": 3.8041,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371319273331442e-05,
+      "loss": 3.8176,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37048067858039e-05,
+      "loss": 3.8029,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369642083829338e-05,
+      "loss": 3.8066,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368803489078286e-05,
+      "loss": 3.8054,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367968170087981e-05,
+      "loss": 3.8036,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367129575336929e-05,
+      "loss": 3.8043,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366290980585877e-05,
+      "loss": 3.8082,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365452385834825e-05,
+      "loss": 3.8081,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3646154289641456e-05,
+      "loss": 3.8025,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3637768342130936e-05,
+      "loss": 3.795,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3629382394620416e-05,
+      "loss": 3.8022,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3620996447109896e-05,
+      "loss": 3.7877,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3612610499599376e-05,
+      "loss": 3.7975,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3604224552088856e-05,
+      "loss": 3.7994,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595838604578336e-05,
+      "loss": 3.7985,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3587452657067816e-05,
+      "loss": 3.82,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3579066709557295e-05,
+      "loss": 3.8066,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570680762046775e-05,
+      "loss": 3.8101,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562294814536255e-05,
+      "loss": 3.8009,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355392524582947e-05,
+      "loss": 3.8044,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354553929831895e-05,
+      "loss": 3.7964,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353715335080843e-05,
+      "loss": 3.8039,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528767403297904e-05,
+      "loss": 3.7966,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3520381455787384e-05,
+      "loss": 3.7955,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511995508276864e-05,
+      "loss": 3.7909,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503609560766344e-05,
+      "loss": 3.7919,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3495223613255824e-05,
+      "loss": 3.7975,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348685404454903e-05,
+      "loss": 3.8055,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347846809703851e-05,
+      "loss": 3.799,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347008214952799e-05,
+      "loss": 3.8064,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346169620201748e-05,
+      "loss": 3.7874,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345332663331069e-05,
+      "loss": 3.7977,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344494068580017e-05,
+      "loss": 3.7916,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343655473828965e-05,
+      "loss": 3.7854,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342816879077913e-05,
+      "loss": 3.7931,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341979922207234e-05,
+      "loss": 3.7797,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341141327456182e-05,
+      "loss": 3.7955,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34030273270513e-05,
+      "loss": 3.7982,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339464137954078e-05,
+      "loss": 3.7986,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338628818963772e-05,
+      "loss": 3.7892,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33779022421272e-05,
+      "loss": 3.8023,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336951629461668e-05,
+      "loss": 3.781,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336113034710616e-05,
+      "loss": 3.7983,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335274439959564e-05,
+      "loss": 3.7871,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334435845208512e-05,
+      "loss": 3.7668,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33359725045746e-05,
+      "loss": 3.8028,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332760293586781e-05,
+      "loss": 3.7853,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331921698835729e-05,
+      "loss": 3.793,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331083104084677e-05,
+      "loss": 3.7798,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330244509333625e-05,
+      "loss": 3.7801,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329405914582573e-05,
+      "loss": 3.7797,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328567319831521e-05,
+      "loss": 3.7836,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327728725080469e-05,
+      "loss": 3.7907,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32689176820979e-05,
+      "loss": 3.7854,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326053173458739e-05,
+      "loss": 3.7979,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325214578707687e-05,
+      "loss": 3.7806,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324375983956635e-05,
+      "loss": 3.7706,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3235390270859556e-05,
+      "loss": 3.7899,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3227004323349036e-05,
+      "loss": 3.7708,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3218618375838516e-05,
+      "loss": 3.7707,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3210232428327996e-05,
+      "loss": 3.7913,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3201862859621205e-05,
+      "loss": 3.7877,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3193493290914414e-05,
+      "loss": 3.776,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3185107343403894e-05,
+      "loss": 3.7757,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3176721395893374e-05,
+      "loss": 3.7667,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3168335448382854e-05,
+      "loss": 3.7769,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315994950087234e-05,
+      "loss": 3.7922,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315156355336182e-05,
+      "loss": 3.7868,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31431776058513e-05,
+      "loss": 3.7883,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313480803714451e-05,
+      "loss": 3.7824,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312642208963399e-05,
+      "loss": 3.7976,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311803614212347e-05,
+      "loss": 3.7737,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310965019461295e-05,
+      "loss": 3.7882,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310126424710243e-05,
+      "loss": 3.7797,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309287829959191e-05,
+      "loss": 3.7639,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308449235208139e-05,
+      "loss": 3.7922,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30761227833746e-05,
+      "loss": 3.7865,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306773683586408e-05,
+      "loss": 3.7835,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305935088835356e-05,
+      "loss": 3.7776,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305096494084304e-05,
+      "loss": 3.7668,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042595372136254e-05,
+      "loss": 3.7724,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034225803429464e-05,
+      "loss": 3.7775,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025839855918944e-05,
+      "loss": 3.785,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3017453908408424e-05,
+      "loss": 3.7728,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3009067960897903e-05,
+      "loss": 3.7809,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000682013387383e-05,
+      "loss": 3.7722,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299231244468059e-05,
+      "loss": 3.7767,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298392649717007e-05,
+      "loss": 3.7707,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297554054965955e-05,
+      "loss": 3.7738,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296715460214903e-05,
+      "loss": 3.7857,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295876865463851e-05,
+      "loss": 3.7898,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295038270712799e-05,
+      "loss": 3.7897,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294199675961747e-05,
+      "loss": 3.7611,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293361081210695e-05,
+      "loss": 3.7802,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292524124340016e-05,
+      "loss": 3.7705,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291687167469338e-05,
+      "loss": 3.7834,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290848572718286e-05,
+      "loss": 3.7722,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290009977967234e-05,
+      "loss": 3.7757,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289171383216181e-05,
+      "loss": 3.7661,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288332788465129e-05,
+      "loss": 3.7814,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287494193714077e-05,
+      "loss": 3.7592,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286655598963025e-05,
+      "loss": 3.7776,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285817004211973e-05,
+      "loss": 3.7624,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849800473412946e-05,
+      "loss": 3.7805,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2841414525902426e-05,
+      "loss": 3.77,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2833028578391906e-05,
+      "loss": 3.7732,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824642630881386e-05,
+      "loss": 3.7745,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2816273062174595e-05,
+      "loss": 3.7735,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807887114664075e-05,
+      "loss": 3.7777,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799501167153555e-05,
+      "loss": 3.7702,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2791115219643035e-05,
+      "loss": 3.7757,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2782729272132515e-05,
+      "loss": 3.7892,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2774343324621995e-05,
+      "loss": 3.7709,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2765957377111475e-05,
+      "loss": 3.7692,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2757587808404684e-05,
+      "loss": 3.7742,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2749201860894164e-05,
+      "loss": 3.7695,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274081591338365e-05,
+      "loss": 3.7629,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273242996587313e-05,
+      "loss": 3.774,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272404401836261e-05,
+      "loss": 3.7591,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271565807085209e-05,
+      "loss": 3.7801,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27072885021453e-05,
+      "loss": 3.7779,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269890255463478e-05,
+      "loss": 3.7733,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269051660712426e-05,
+      "loss": 3.7647,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268213065961374e-05,
+      "loss": 3.7636,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267374471210322e-05,
+      "loss": 3.7679,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26653587645927e-05,
+      "loss": 3.7768,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265697281708217e-05,
+      "loss": 3.7715,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264858686957165e-05,
+      "loss": 3.7762,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264021730086487e-05,
+      "loss": 3.7525,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263183135335435e-05,
+      "loss": 3.776,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262344540584383e-05,
+      "loss": 3.7684,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2615075837137044e-05,
+      "loss": 3.7763,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606689889626524e-05,
+      "loss": 3.7598,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2598303942116e-05,
+      "loss": 3.7785,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258991799460548e-05,
+      "loss": 3.7706,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258154842589869e-05,
+      "loss": 3.7804,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257316247838817e-05,
+      "loss": 3.7641,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564776530877646e-05,
+      "loss": 3.7756,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2556390583367126e-05,
+      "loss": 3.7558,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2548004635856606e-05,
+      "loss": 3.7694,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2539618688346086e-05,
+      "loss": 3.7729,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25312491196393e-05,
+      "loss": 3.7642,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252286317212878e-05,
+      "loss": 3.7684,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251447722461826e-05,
+      "loss": 3.7614,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250609127710774e-05,
+      "loss": 3.7702,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.891617774963379,
+      "eval_runtime": 303.681,
+      "eval_samples_per_second": 1256.552,
+      "eval_steps_per_second": 39.268,
+      "step": 457920
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.169689251716915e+17,
+  "trial_name": null,
+  "trial_params": null
+}