diff --git "a/checkpoint-915840/trainer_state.json" "b/checkpoint-915840/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-915840/trainer_state.json"
@@ -0,0 +1,10849 @@
+{
+  "best_metric": 3.849750280380249,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl2/transformer/2/checkpoints/checkpoint-839520",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 915840,
+  "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": 10.9836,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8172,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1904,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9873,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.827,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7261,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6163,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5478,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4726,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.412,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3734,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3286,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2826,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 5.2333,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 5.2069,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 5.1604,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1343,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.1101,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0768,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0456,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.0296,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9974,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9807,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9544,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9469,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.9125,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8982,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8691,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8618,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8367,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 4.8312,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740117521192533e-05,
+      "loss": 4.8185,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731731573682013e-05,
+      "loss": 4.8007,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7903,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7752,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.761,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.757,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7261,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7164,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6952,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6981,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6834,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6846,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6621,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6564,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6526,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6467,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6325,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6173,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 4.6071,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6255,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.6142,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5923,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5784,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5873,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5644,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5751,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5425,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5655,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5495,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5293,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 4.5227,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5342,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4947,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.5078,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5164,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.5005,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.496,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4763,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4754,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4703,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4878,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.463,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4726,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4679,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4669,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4479,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4468,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4423,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4389,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4327,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4249,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4275,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.4254,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.4049,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.4122,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4116,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.4129,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3978,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 4.3971,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3872,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3816,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3838,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3866,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3787,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3861,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3761,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3614,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3604,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3672,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3649,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3638,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3505,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3409,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.355,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3334,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3381,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3349,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3339,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3247,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.3256,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3179,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.325,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3191,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3115,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3216,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.3277,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.3146,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.3022,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.299,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.298,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.3032,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.2983,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.2871,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2974,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2875,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2855,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.2823,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2843,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.2756,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2864,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.2842,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.283,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.2604,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2746,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.2688,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.2727,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.261,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884314871364151e-05,
+      "loss": 4.2695,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.2593,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2759,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.2471,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.2634,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.2486,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2463,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.2489,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.2351,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.2402,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.2456,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.2351,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.202381610870361,
+      "eval_runtime": 313.8478,
+      "eval_samples_per_second": 1215.847,
+      "eval_steps_per_second": 37.996,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.2299,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.2254,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.2407,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.2307,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.24,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.2184,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.2201,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.2078,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.2168,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.2153,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865873976242873e-05,
+      "loss": 4.2126,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.2169,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.1961,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.2095,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.1935,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.194,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.2005,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.193,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.1962,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.2109,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1859,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.1908,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.1892,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1938,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.186,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.1872,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.1718,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.1782,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1663,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1778,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1739,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1763,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.1797,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.1678,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.1742,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1693,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844078702117387e-05,
+      "loss": 4.16,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843240107366335e-05,
+      "loss": 4.1529,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1439,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1521,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.1508,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.1599,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1454,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1461,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.1521,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.1468,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1459,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1334,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834020478745882e-05,
+      "loss": 4.1309,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.1536,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1517,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1352,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830667737622047e-05,
+      "loss": 4.1361,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298291428709944e-05,
+      "loss": 4.1386,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289905481199424e-05,
+      "loss": 4.1225,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.1422,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273133586178384e-05,
+      "loss": 4.1148,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.142,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.1351,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824799212245056e-05,
+      "loss": 4.1173,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.1162,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.1304,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.0958,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.1146,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820607876370169e-05,
+      "loss": 4.1278,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819769281619117e-05,
+      "loss": 4.1195,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818930686868065e-05,
+      "loss": 4.1115,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818092092117013e-05,
+      "loss": 4.0986,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817255135246334e-05,
+      "loss": 4.1038,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816416540495282e-05,
+      "loss": 4.1074,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155779457442305e-05,
+      "loss": 4.1149,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.1037,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139023941224994e-05,
+      "loss": 4.1115,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130637993714474e-05,
+      "loss": 4.1151,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122252046203954e-05,
+      "loss": 4.1135,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113866098693434e-05,
+      "loss": 4.097,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810549652998664e-05,
+      "loss": 4.1048,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809711058247612e-05,
+      "loss": 4.1036,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887246349656e-05,
+      "loss": 4.0989,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808033868745508e-05,
+      "loss": 4.0994,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807195273994456e-05,
+      "loss": 4.0928,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806358317123777e-05,
+      "loss": 4.1019,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805519722372726e-05,
+      "loss": 4.0981,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804681127621674e-05,
+      "loss": 4.081,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803842532870622e-05,
+      "loss": 4.0901,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803005575999943e-05,
+      "loss": 4.0931,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802166981248891e-05,
+      "loss": 4.0978,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801328386497839e-05,
+      "loss": 4.09,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004914296271596e-05,
+      "loss": 4.0855,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996528348761076e-05,
+      "loss": 4.0804,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988142401250556e-05,
+      "loss": 4.0759,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979756453740036e-05,
+      "loss": 4.0774,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971370506229516e-05,
+      "loss": 4.0886,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7962984558718996e-05,
+      "loss": 4.084,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954598611208476e-05,
+      "loss": 4.0909,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946212663697956e-05,
+      "loss": 4.0831,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793784309499117e-05,
+      "loss": 4.0735,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792945714748065e-05,
+      "loss": 4.0727,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792107119997013e-05,
+      "loss": 4.0796,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912685252459605e-05,
+      "loss": 4.0854,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790431568375282e-05,
+      "loss": 4.0822,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78959297362423e-05,
+      "loss": 4.0735,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887543788731774e-05,
+      "loss": 4.0636,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879157841221254e-05,
+      "loss": 4.0832,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787078827251447e-05,
+      "loss": 4.0619,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786241870380768e-05,
+      "loss": 4.0694,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854032756297166e-05,
+      "loss": 4.0694,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845646808786646e-05,
+      "loss": 4.0675,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837260861276126e-05,
+      "loss": 4.0566,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828874913765606e-05,
+      "loss": 4.0667,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782048896625508e-05,
+      "loss": 4.0566,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781210301874456e-05,
+      "loss": 4.064,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803733450037775e-05,
+      "loss": 4.0641,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779534750252725e-05,
+      "loss": 4.0547,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778696155501673e-05,
+      "loss": 4.075,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777857560750621e-05,
+      "loss": 4.0744,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777018965999569e-05,
+      "loss": 4.0648,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776180371248517e-05,
+      "loss": 4.0559,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775341776497465e-05,
+      "loss": 4.0499,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7745031817464134e-05,
+      "loss": 4.056,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773666224875734e-05,
+      "loss": 4.0533,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772827630124682e-05,
+      "loss": 4.0622,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77198903537363e-05,
+      "loss": 4.0431,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771150440622578e-05,
+      "loss": 4.053,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770313483751899e-05,
+      "loss": 4.0557,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769474889000847e-05,
+      "loss": 4.0552,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768636294249795e-05,
+      "loss": 4.0445,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767797699498743e-05,
+      "loss": 4.0491,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766960742628064e-05,
+      "loss": 4.049,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766122147877012e-05,
+      "loss": 4.0527,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528355312596e-05,
+      "loss": 4.0554,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764444958374909e-05,
+      "loss": 4.0555,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76360800150423e-05,
+      "loss": 4.0334,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762769406753178e-05,
+      "loss": 4.0472,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761930812002126e-05,
+      "loss": 4.0505,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761092217251074e-05,
+      "loss": 4.0471,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602552603803946e-05,
+      "loss": 4.0416,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594166656293426e-05,
+      "loss": 4.0495,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585797087586635e-05,
+      "loss": 4.0432,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577411140076115e-05,
+      "loss": 4.0593,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569025192565595e-05,
+      "loss": 4.0338,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560639245055075e-05,
+      "loss": 4.0473,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552253297544555e-05,
+      "loss": 4.0376,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543867350034035e-05,
+      "loss": 4.0328,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753549778132725e-05,
+      "loss": 4.0392,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752711183381673e-05,
+      "loss": 4.0291,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751872588630621e-05,
+      "loss": 4.0328,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751033993879569e-05,
+      "loss": 4.0343,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750195399128517e-05,
+      "loss": 4.0288,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.033535957336426,
+      "eval_runtime": 333.1084,
+      "eval_samples_per_second": 1145.546,
+      "eval_steps_per_second": 35.799,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749356804377465e-05,
+      "loss": 4.0263,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748518209626413e-05,
+      "loss": 4.0209,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747679614875361e-05,
+      "loss": 4.0395,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468410201243084e-05,
+      "loss": 4.029,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460024253732563e-05,
+      "loss": 4.0438,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451638306222043e-05,
+      "loss": 4.0187,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744325235871152e-05,
+      "loss": 4.0257,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434866411201e-05,
+      "loss": 4.009,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742649684249422e-05,
+      "loss": 4.027,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74181108949837e-05,
+      "loss": 4.0226,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740972494747318e-05,
+      "loss": 4.0177,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740133899996266e-05,
+      "loss": 4.0293,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739295305245214e-05,
+      "loss": 4.0042,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738456710494162e-05,
+      "loss": 4.023,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73761811574311e-05,
+      "loss": 4.0054,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736779520992058e-05,
+      "loss": 4.0063,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735942564121379e-05,
+      "loss": 4.0069,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735103969370327e-05,
+      "loss": 4.0162,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734265374619275e-05,
+      "loss": 4.0091,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733426779868223e-05,
+      "loss": 4.0287,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325898229975444e-05,
+      "loss": 4.0091,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317512282464924e-05,
+      "loss": 4.0116,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309126334954404e-05,
+      "loss": 4.0094,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300740387443884e-05,
+      "loss": 4.0171,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729237081873709e-05,
+      "loss": 4.0061,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728398487122657e-05,
+      "loss": 4.0128,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727559892371605e-05,
+      "loss": 4.0007,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726721297620553e-05,
+      "loss": 4.0044,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725884340749874e-05,
+      "loss": 3.9914,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725045745998822e-05,
+      "loss": 4.006,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72420715124777e-05,
+      "loss": 4.0053,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723368556496718e-05,
+      "loss": 4.0058,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72253159962604e-05,
+      "loss": 4.01,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721693004874988e-05,
+      "loss": 4.0002,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720854410123936e-05,
+      "loss": 3.9994,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720015815372884e-05,
+      "loss": 4.0053,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191788585022046e-05,
+      "loss": 3.9966,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183402637511526e-05,
+      "loss": 3.9875,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175016690001006e-05,
+      "loss": 3.9792,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166630742490486e-05,
+      "loss": 3.9872,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158261173783695e-05,
+      "loss": 3.9908,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149875226273175e-05,
+      "loss": 3.9969,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141489278762655e-05,
+      "loss": 3.9909,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133103331252135e-05,
+      "loss": 3.982,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712473376254535e-05,
+      "loss": 3.9965,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711634781503483e-05,
+      "loss": 3.9894,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710796186752431e-05,
+      "loss": 3.9846,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709957592001379e-05,
+      "loss": 3.9791,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091206351307e-05,
+      "loss": 3.9728,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708283678260021e-05,
+      "loss": 3.9978,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707445083508969e-05,
+      "loss": 3.9962,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706606488757917e-05,
+      "loss": 3.9808,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705767894006865e-05,
+      "loss": 3.9818,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704930937136186e-05,
+      "loss": 3.9883,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704092342385134e-05,
+      "loss": 3.9664,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703253747634082e-05,
+      "loss": 3.9882,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024151528830305e-05,
+      "loss": 3.9657,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015765581319785e-05,
+      "loss": 3.9858,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007379633809265e-05,
+      "loss": 3.9909,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998993686298745e-05,
+      "loss": 3.9712,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990624117591954e-05,
+      "loss": 3.9639,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982238170081434e-05,
+      "loss": 3.9838,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973852222570914e-05,
+      "loss": 3.9531,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965466275060394e-05,
+      "loss": 3.9692,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69570967063536e-05,
+      "loss": 3.9804,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694871075884308e-05,
+      "loss": 3.9755,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694032481133256e-05,
+      "loss": 3.9695,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693193886382204e-05,
+      "loss": 3.955,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692358567391899e-05,
+      "loss": 3.9579,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.9621,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690681377889795e-05,
+      "loss": 3.9739,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689842783138743e-05,
+      "loss": 3.9614,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689004188387691e-05,
+      "loss": 3.9733,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688165593636639e-05,
+      "loss": 3.9738,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687326998885587e-05,
+      "loss": 3.9739,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686488404134534e-05,
+      "loss": 3.9578,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856514472638556e-05,
+      "loss": 3.9641,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848128525128036e-05,
+      "loss": 3.9692,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683974257761751e-05,
+      "loss": 3.96,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831373008910725e-05,
+      "loss": 3.9634,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822987061400205e-05,
+      "loss": 3.9605,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681460111388969e-05,
+      "loss": 3.9619,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806215166379165e-05,
+      "loss": 3.9654,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797829218868645e-05,
+      "loss": 3.9455,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789443271358125e-05,
+      "loss": 3.9596,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678107370265134e-05,
+      "loss": 3.9557,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772687755140814e-05,
+      "loss": 3.9647,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764301807630294e-05,
+      "loss": 3.9542,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755915860119774e-05,
+      "loss": 3.955,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747529912609254e-05,
+      "loss": 3.9514,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739143965098734e-05,
+      "loss": 3.9469,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730758017588214e-05,
+      "loss": 3.9435,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722372070077694e-05,
+      "loss": 3.9562,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671401888017464e-05,
+      "loss": 3.9649,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670563293266412e-05,
+      "loss": 3.9593,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66972469851536e-05,
+      "loss": 3.9521,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668886103764308e-05,
+      "loss": 3.9453,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668047509013256e-05,
+      "loss": 3.9472,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667208914262204e-05,
+      "loss": 3.9507,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666370319511152e-05,
+      "loss": 3.96,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655317247601e-05,
+      "loss": 3.9545,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664694767889421e-05,
+      "loss": 3.9485,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663856173138369e-05,
+      "loss": 3.9383,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663017578387317e-05,
+      "loss": 3.9593,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662178983636265e-05,
+      "loss": 3.9384,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661343664645959e-05,
+      "loss": 3.9485,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660505069894907e-05,
+      "loss": 3.9434,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659666475143855e-05,
+      "loss": 3.9471,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658827880392803e-05,
+      "loss": 3.936,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657989285641751e-05,
+      "loss": 3.947,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657150690890699e-05,
+      "loss": 3.9379,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65631373402002e-05,
+      "loss": 3.9379,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655475139268968e-05,
+      "loss": 3.948,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654636544517916e-05,
+      "loss": 3.9314,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653797949766864e-05,
+      "loss": 3.955,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652959355015812e-05,
+      "loss": 3.9541,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652122398145134e-05,
+      "loss": 3.9441,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651283803394082e-05,
+      "loss": 3.9393,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65044520864303e-05,
+      "loss": 3.9341,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649606613891978e-05,
+      "loss": 3.9343,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648768019140926e-05,
+      "loss": 3.9357,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479310622702466e-05,
+      "loss": 3.9446,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470924675191946e-05,
+      "loss": 3.9268,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462538727681426e-05,
+      "loss": 3.9376,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454152780170906e-05,
+      "loss": 3.941,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445766832660386e-05,
+      "loss": 3.9386,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437380885149866e-05,
+      "loss": 3.9311,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429011316443075e-05,
+      "loss": 3.934,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420625368932555e-05,
+      "loss": 3.9375,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412239421422035e-05,
+      "loss": 3.9368,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640385347391152e-05,
+      "loss": 3.9395,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395467526401e-05,
+      "loss": 3.9424,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387081578890475e-05,
+      "loss": 3.9207,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378695631379955e-05,
+      "loss": 3.9361,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370309683869435e-05,
+      "loss": 3.9392,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636194011516265e-05,
+      "loss": 3.9324,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353554167652124e-05,
+      "loss": 3.932,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345168220141604e-05,
+      "loss": 3.9412,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336782272631084e-05,
+      "loss": 3.9329,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328396325120564e-05,
+      "loss": 3.9472,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632002675641377e-05,
+      "loss": 3.9236,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631164080890326e-05,
+      "loss": 3.9328,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630325486139274e-05,
+      "loss": 3.9303,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629486891388222e-05,
+      "loss": 3.9244,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628649934517543e-05,
+      "loss": 3.9286,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627811339766491e-05,
+      "loss": 3.9234,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626972745015439e-05,
+      "loss": 3.9227,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626134150264387e-05,
+      "loss": 3.926,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625295555513335e-05,
+      "loss": 3.9259,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9588358402252197,
+      "eval_runtime": 307.1405,
+      "eval_samples_per_second": 1242.399,
+      "eval_steps_per_second": 38.826,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624456960762283e-05,
+      "loss": 3.9232,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623620003891604e-05,
+      "loss": 3.9155,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622781409140552e-05,
+      "loss": 3.9304,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219428143895e-05,
+      "loss": 3.9217,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621104219638448e-05,
+      "loss": 3.9418,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620265624887396e-05,
+      "loss": 3.9162,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619428668016717e-05,
+      "loss": 3.92,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618590073265665e-05,
+      "loss": 3.9041,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617751478514613e-05,
+      "loss": 3.9227,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616912883763561e-05,
+      "loss": 3.9219,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616074289012509e-05,
+      "loss": 3.913,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61523733214183e-05,
+      "loss": 3.9201,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614398737390778e-05,
+      "loss": 3.9049,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613560142639726e-05,
+      "loss": 3.9198,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612721547888674e-05,
+      "loss": 3.9041,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611884591017995e-05,
+      "loss": 3.904,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611045996266943e-05,
+      "loss": 3.9068,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610207401515891e-05,
+      "loss": 3.9127,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609368806764839e-05,
+      "loss": 3.9102,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085318498941607e-05,
+      "loss": 3.9282,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076932551431087e-05,
+      "loss": 3.9099,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068546603920566e-05,
+      "loss": 3.9169,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060160656410046e-05,
+      "loss": 3.9089,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051774708899526e-05,
+      "loss": 3.9157,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043388761389006e-05,
+      "loss": 3.9064,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603500281387848e-05,
+      "loss": 3.9134,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602661686636796e-05,
+      "loss": 3.9061,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018247297661175e-05,
+      "loss": 3.9027,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009877728954384e-05,
+      "loss": 3.8998,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001491781443864e-05,
+      "loss": 3.9055,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993105833933344e-05,
+      "loss": 3.9106,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598471988642283e-05,
+      "loss": 3.9077,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597633393891231e-05,
+      "loss": 3.9173,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967947991401784e-05,
+      "loss": 3.9062,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959562043891264e-05,
+      "loss": 3.9019,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595119247518448e-05,
+      "loss": 3.907,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594280652767395e-05,
+      "loss": 3.9064,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593442058016343e-05,
+      "loss": 3.8888,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592603463265291e-05,
+      "loss": 3.8866,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591766506394613e-05,
+      "loss": 3.8918,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59092791164356e-05,
+      "loss": 3.9007,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590089316892508e-05,
+      "loss": 3.9009,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589250722141457e-05,
+      "loss": 3.8994,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884137652707785e-05,
+      "loss": 3.887,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587575170519726e-05,
+      "loss": 3.9031,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586736575768674e-05,
+      "loss": 3.8972,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585897981017622e-05,
+      "loss": 3.8924,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58505938626657e-05,
+      "loss": 3.8849,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584222429395891e-05,
+      "loss": 3.8765,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583383834644839e-05,
+      "loss": 3.9132,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582545239893787e-05,
+      "loss": 3.9012,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581706645142735e-05,
+      "loss": 3.8909,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808696882720556e-05,
+      "loss": 3.8954,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800310935210036e-05,
+      "loss": 3.8924,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579192498769952e-05,
+      "loss": 3.8775,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783539040189e-05,
+      "loss": 3.8955,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577515309267848e-05,
+      "loss": 3.8813,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576678352397169e-05,
+      "loss": 3.8924,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575839757646117e-05,
+      "loss": 3.9049,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575001162895065e-05,
+      "loss": 3.8836,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574162568144013e-05,
+      "loss": 3.8716,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573325611273334e-05,
+      "loss": 3.9009,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572487016522282e-05,
+      "loss": 3.8669,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57164842177123e-05,
+      "loss": 3.8775,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570809827020178e-05,
+      "loss": 3.8955,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569971232269126e-05,
+      "loss": 3.8882,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569132637518074e-05,
+      "loss": 3.8812,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682956806473956e-05,
+      "loss": 3.873,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674570858963436e-05,
+      "loss": 3.8667,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666184911452916e-05,
+      "loss": 3.876,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657798963942396e-05,
+      "loss": 3.8892,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649429395235605e-05,
+      "loss": 3.873,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641043447725085e-05,
+      "loss": 3.8901,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632657500214565e-05,
+      "loss": 3.8832,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624271552704045e-05,
+      "loss": 3.892,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615885605193525e-05,
+      "loss": 3.8774,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607516036486734e-05,
+      "loss": 3.876,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599130088976214e-05,
+      "loss": 3.886,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590744141465694e-05,
+      "loss": 3.8712,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582358193955174e-05,
+      "loss": 3.886,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557398862524839e-05,
+      "loss": 3.8771,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556560267773787e-05,
+      "loss": 3.8779,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555721673022735e-05,
+      "loss": 3.8793,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554883078271683e-05,
+      "loss": 3.8659,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554044483520631e-05,
+      "loss": 3.872,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553207526649952e-05,
+      "loss": 3.8764,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523689318989e-05,
+      "loss": 3.8804,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551530337147848e-05,
+      "loss": 3.8693,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550691742396796e-05,
+      "loss": 3.8739,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549853147645744e-05,
+      "loss": 3.8676,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549016190775065e-05,
+      "loss": 3.8636,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548177596024013e-05,
+      "loss": 3.8647,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547339001272961e-05,
+      "loss": 3.8713,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546500406521909e-05,
+      "loss": 3.8818,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54566344965123e-05,
+      "loss": 3.8828,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544824854900178e-05,
+      "loss": 3.8727,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543986260149126e-05,
+      "loss": 3.8633,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431476653980736e-05,
+      "loss": 3.8666,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542310708527395e-05,
+      "loss": 3.872,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541472113776343e-05,
+      "loss": 3.8763,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540633519025291e-05,
+      "loss": 3.8755,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397949242742385e-05,
+      "loss": 3.8707,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53895796740356e-05,
+      "loss": 3.859,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538119372652508e-05,
+      "loss": 3.8823,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537280777901456e-05,
+      "loss": 3.8554,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536442183150404e-05,
+      "loss": 3.8734,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535605226279726e-05,
+      "loss": 3.8613,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534766631528674e-05,
+      "loss": 3.8682,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533928036777621e-05,
+      "loss": 3.8572,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533089442026569e-05,
+      "loss": 3.8684,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322524851558906e-05,
+      "loss": 3.8616,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314138904048386e-05,
+      "loss": 3.8622,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530575295653786e-05,
+      "loss": 3.8683,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529736700902734e-05,
+      "loss": 3.8574,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5288997440320555e-05,
+      "loss": 3.8741,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280611492810035e-05,
+      "loss": 3.8755,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272225545299515e-05,
+      "loss": 3.8701,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263839597788995e-05,
+      "loss": 3.8607,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255453650278475e-05,
+      "loss": 3.8593,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247084081571684e-05,
+      "loss": 3.8573,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238698134061164e-05,
+      "loss": 3.86,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230312186550644e-05,
+      "loss": 3.8684,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221926239040124e-05,
+      "loss": 3.8466,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521355667033333e-05,
+      "loss": 3.8621,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520517072282281e-05,
+      "loss": 3.861,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519678477531229e-05,
+      "loss": 3.8664,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518839882780177e-05,
+      "loss": 3.8542,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518002925909499e-05,
+      "loss": 3.86,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517164331158447e-05,
+      "loss": 3.8571,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516325736407395e-05,
+      "loss": 3.8698,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515487141656343e-05,
+      "loss": 3.8621,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514650184785664e-05,
+      "loss": 3.87,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513811590034612e-05,
+      "loss": 3.8458,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51297299528356e-05,
+      "loss": 3.8624,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512134400532508e-05,
+      "loss": 3.8653,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112974436618287e-05,
+      "loss": 3.8552,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104588489107766e-05,
+      "loss": 3.8561,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096202541597246e-05,
+      "loss": 3.8689,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087816594086726e-05,
+      "loss": 3.8613,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507944702537994e-05,
+      "loss": 3.8762,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507106107786942e-05,
+      "loss": 3.8483,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50626751303589e-05,
+      "loss": 3.8629,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505428918284838e-05,
+      "loss": 3.8526,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504591961414159e-05,
+      "loss": 3.8536,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503753366663107e-05,
+      "loss": 3.8576,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502914771912055e-05,
+      "loss": 3.8505,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502076177161003e-05,
+      "loss": 3.8516,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501239220290324e-05,
+      "loss": 3.8504,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500400625539272e-05,
+      "loss": 3.8543,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9177722930908203,
+      "eval_runtime": 307.9626,
+      "eval_samples_per_second": 1239.082,
+      "eval_steps_per_second": 38.722,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49956203078822e-05,
+      "loss": 3.8468,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498723436037168e-05,
+      "loss": 3.8432,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497884841286116e-05,
+      "loss": 3.8593,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497046246535065e-05,
+      "loss": 3.8503,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496207651784013e-05,
+      "loss": 3.8651,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495369057032961e-05,
+      "loss": 3.8512,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945304622819087e-05,
+      "loss": 3.8491,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936918675308567e-05,
+      "loss": 3.8352,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928549106601776e-05,
+      "loss": 3.8504,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920163159091256e-05,
+      "loss": 3.8486,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911777211580736e-05,
+      "loss": 3.8472,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903407642873945e-05,
+      "loss": 3.8485,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895021695363425e-05,
+      "loss": 3.8343,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886635747852905e-05,
+      "loss": 3.8515,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878249800342385e-05,
+      "loss": 3.8346,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869863852831864e-05,
+      "loss": 3.8341,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861477905321344e-05,
+      "loss": 3.836,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853091957810824e-05,
+      "loss": 3.8475,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484472238910404e-05,
+      "loss": 3.8399,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483633644159352e-05,
+      "loss": 3.8573,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827950494082993e-05,
+      "loss": 3.8435,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481956454657247e-05,
+      "loss": 3.8516,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481119497786569e-05,
+      "loss": 3.8399,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480280903035516e-05,
+      "loss": 3.8445,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479442308284464e-05,
+      "loss": 3.8387,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478603713533412e-05,
+      "loss": 3.8482,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47776511878236e-05,
+      "loss": 3.8381,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476926524031308e-05,
+      "loss": 3.8375,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47608956716063e-05,
+      "loss": 3.8304,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475250972409578e-05,
+      "loss": 3.8348,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474412377658526e-05,
+      "loss": 3.8466,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473573782907474e-05,
+      "loss": 3.8431,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472736826036795e-05,
+      "loss": 3.8463,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471898231285743e-05,
+      "loss": 3.8354,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471059636534691e-05,
+      "loss": 3.8369,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470221041783639e-05,
+      "loss": 3.8463,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693840849129596e-05,
+      "loss": 3.8435,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685454901619076e-05,
+      "loss": 3.8194,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677068954108556e-05,
+      "loss": 3.8252,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668683006598036e-05,
+      "loss": 3.8239,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466031343789125e-05,
+      "loss": 3.8306,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465192749038073e-05,
+      "loss": 3.8391,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464354154287021e-05,
+      "loss": 3.833,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463515559535969e-05,
+      "loss": 3.8288,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46267860266529e-05,
+      "loss": 3.8341,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461840007914238e-05,
+      "loss": 3.83,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461001413163186e-05,
+      "loss": 3.8322,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460162818412134e-05,
+      "loss": 3.8236,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459325861541455e-05,
+      "loss": 3.8062,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458487266790403e-05,
+      "loss": 3.8503,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457648672039351e-05,
+      "loss": 3.8351,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456810077288299e-05,
+      "loss": 3.8332,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559731204176205e-05,
+      "loss": 3.8301,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551361635469415e-05,
+      "loss": 3.8264,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542975687958895e-05,
+      "loss": 3.8158,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534589740448374e-05,
+      "loss": 3.8301,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526203792937854e-05,
+      "loss": 3.8229,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517817845427334e-05,
+      "loss": 3.8256,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509431897916814e-05,
+      "loss": 3.8418,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501045950406294e-05,
+      "loss": 3.821,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492660002895774e-05,
+      "loss": 3.8126,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448429043418898e-05,
+      "loss": 3.8363,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447590448667846e-05,
+      "loss": 3.805,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446751853916794e-05,
+      "loss": 3.8176,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445913259165742e-05,
+      "loss": 3.8292,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445076302295064e-05,
+      "loss": 3.826,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444237707544012e-05,
+      "loss": 3.8187,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44339911279296e-05,
+      "loss": 3.8137,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442563793802654e-05,
+      "loss": 3.8019,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441725199051602e-05,
+      "loss": 3.8145,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44088660430055e-05,
+      "loss": 3.826,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440048009549498e-05,
+      "loss": 3.8138,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439209414798446e-05,
+      "loss": 3.8245,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438370820047394e-05,
+      "loss": 3.8255,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437532225296342e-05,
+      "loss": 3.8334,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669363054529e-05,
+      "loss": 3.8156,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435856673674611e-05,
+      "loss": 3.8136,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435018078923559e-05,
+      "loss": 3.8243,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434179484172507e-05,
+      "loss": 3.8092,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433340889421455e-05,
+      "loss": 3.8289,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432503932550776e-05,
+      "loss": 3.8177,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431665337799724e-05,
+      "loss": 3.816,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430826743048672e-05,
+      "loss": 3.8226,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42998814829762e-05,
+      "loss": 3.8041,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429149553546568e-05,
+      "loss": 3.8091,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428312596675889e-05,
+      "loss": 3.8156,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427474001924837e-05,
+      "loss": 3.8224,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426635407173785e-05,
+      "loss": 3.8066,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425796812422733e-05,
+      "loss": 3.8141,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249598555520546e-05,
+      "loss": 3.808,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241212608010026e-05,
+      "loss": 3.8078,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232826660499506e-05,
+      "loss": 3.8015,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224440712988986e-05,
+      "loss": 3.8152,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216071144282195e-05,
+      "loss": 3.8224,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207701575575404e-05,
+      "loss": 3.8195,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199315628064884e-05,
+      "loss": 3.8178,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190929680554364e-05,
+      "loss": 3.8042,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182543733043844e-05,
+      "loss": 3.8076,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174157785533324e-05,
+      "loss": 3.8119,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165771838022804e-05,
+      "loss": 3.8189,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157385890512284e-05,
+      "loss": 3.8173,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414899994300177e-05,
+      "loss": 3.8083,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414063037429498e-05,
+      "loss": 3.8042,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413224442678446e-05,
+      "loss": 3.8222,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412385847927394e-05,
+      "loss": 3.8,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411547253176342e-05,
+      "loss": 3.814,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410710296305663e-05,
+      "loss": 3.8042,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409871701554611e-05,
+      "loss": 3.8123,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409034744683932e-05,
+      "loss": 3.7985,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40819614993288e-05,
+      "loss": 3.8114,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407357555181828e-05,
+      "loss": 3.8021,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406518960430776e-05,
+      "loss": 3.8073,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405680365679724e-05,
+      "loss": 3.8107,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404841770928672e-05,
+      "loss": 3.802,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40400317617762e-05,
+      "loss": 3.8138,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403164581426568e-05,
+      "loss": 3.8199,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4023276245558894e-05,
+      "loss": 3.8101,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401489029804837e-05,
+      "loss": 3.8098,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400650435053785e-05,
+      "loss": 3.8001,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399811840302733e-05,
+      "loss": 3.8033,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398974883432054e-05,
+      "loss": 3.8015,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981362886810016e-05,
+      "loss": 3.8164,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972976939299496e-05,
+      "loss": 3.7876,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964590991788976e-05,
+      "loss": 3.8053,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395622142308219e-05,
+      "loss": 3.8025,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394783547557167e-05,
+      "loss": 3.8099,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393944952806115e-05,
+      "loss": 3.8028,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393106358055063e-05,
+      "loss": 3.8047,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392269401184384e-05,
+      "loss": 3.7992,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391430806433332e-05,
+      "loss": 3.8113,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39059221168228e-05,
+      "loss": 3.8096,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389753616931228e-05,
+      "loss": 3.8129,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388916660060549e-05,
+      "loss": 3.7926,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388078065309497e-05,
+      "loss": 3.8066,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387239470558445e-05,
+      "loss": 3.8102,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386400875807393e-05,
+      "loss": 3.8014,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855639189367145e-05,
+      "loss": 3.8021,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847253241856625e-05,
+      "loss": 3.8112,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838867294346105e-05,
+      "loss": 3.8049,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830481346835585e-05,
+      "loss": 3.8245,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822111778128794e-05,
+      "loss": 3.795,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813725830618274e-05,
+      "loss": 3.8064,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805339883107754e-05,
+      "loss": 3.7986,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796953935597234e-05,
+      "loss": 3.7994,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378858436689044e-05,
+      "loss": 3.8051,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378019841937992e-05,
+      "loss": 3.7958,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37718124718694e-05,
+      "loss": 3.8019,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376342652435888e-05,
+      "loss": 3.7869,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37550569556521e-05,
+      "loss": 3.804,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.892151355743408,
+      "eval_runtime": 335.6919,
+      "eval_samples_per_second": 1136.73,
+      "eval_steps_per_second": 35.524,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374667100814158e-05,
+      "loss": 3.7989,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373828506063106e-05,
+      "loss": 3.7905,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372989911312054e-05,
+      "loss": 3.804,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372152954441375e-05,
+      "loss": 3.8023,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371314359690323e-05,
+      "loss": 3.809,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370475764939271e-05,
+      "loss": 3.7948,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369637170188219e-05,
+      "loss": 3.7959,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36880021331754e-05,
+      "loss": 3.7832,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367961618566488e-05,
+      "loss": 3.7971,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367123023815436e-05,
+      "loss": 3.7961,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366284429064384e-05,
+      "loss": 3.798,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365447472193705e-05,
+      "loss": 3.7907,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364610515323026e-05,
+      "loss": 3.7885,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363771920571974e-05,
+      "loss": 3.7922,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362933325820922e-05,
+      "loss": 3.7841,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36209473106987e-05,
+      "loss": 3.7801,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361256136318818e-05,
+      "loss": 3.7836,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360417541567766e-05,
+      "loss": 3.7922,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359578946816714e-05,
+      "loss": 3.7879,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358740352065662e-05,
+      "loss": 3.8041,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357903395194983e-05,
+      "loss": 3.7916,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357064800443931e-05,
+      "loss": 3.8001,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356226205692879e-05,
+      "loss": 3.7865,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355387610941827e-05,
+      "loss": 3.7891,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545506540711486e-05,
+      "loss": 3.7864,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3537120593200966e-05,
+      "loss": 3.7993,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528734645690446e-05,
+      "loss": 3.7877,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520348698179926e-05,
+      "loss": 3.785,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511979129473135e-05,
+      "loss": 3.781,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503593181962615e-05,
+      "loss": 3.7868,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3495207234452095e-05,
+      "loss": 3.7906,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486821286941575e-05,
+      "loss": 3.7933,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478451718234784e-05,
+      "loss": 3.795,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3470065770724264e-05,
+      "loss": 3.7888,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461679823213744e-05,
+      "loss": 3.786,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3453293875703224e-05,
+      "loss": 3.7947,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344492430699644e-05,
+      "loss": 3.7909,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343653835948592e-05,
+      "loss": 3.7752,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34281524119754e-05,
+      "loss": 3.7738,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341976646446488e-05,
+      "loss": 3.7725,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341139689575809e-05,
+      "loss": 3.7772,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340301094824757e-05,
+      "loss": 3.7908,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339462500073705e-05,
+      "loss": 3.7838,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338623905322653e-05,
+      "loss": 3.7767,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337786948451974e-05,
+      "loss": 3.7849,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336948353700922e-05,
+      "loss": 3.7797,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33610975894987e-05,
+      "loss": 3.7828,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335271164198818e-05,
+      "loss": 3.773,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344342073281393e-05,
+      "loss": 3.7567,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3335956125770873e-05,
+      "loss": 3.8024,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327570178260353e-05,
+      "loss": 3.7835,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331918423074983e-05,
+      "loss": 3.7881,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331081466204304e-05,
+      "loss": 3.7767,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330244509333625e-05,
+      "loss": 3.7826,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329405914582573e-05,
+      "loss": 3.7657,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328567319831521e-05,
+      "loss": 3.7772,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327728725080469e-05,
+      "loss": 3.7772,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326890130329417e-05,
+      "loss": 3.7761,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326051535578365e-05,
+      "loss": 3.7959,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325212940827313e-05,
+      "loss": 3.7697,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324374346076261e-05,
+      "loss": 3.7613,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323537389205583e-05,
+      "loss": 3.7898,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322698794454531e-05,
+      "loss": 3.7573,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321860199703479e-05,
+      "loss": 3.7668,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321021604952427e-05,
+      "loss": 3.781,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201846480817476e-05,
+      "loss": 3.7807,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193460533306956e-05,
+      "loss": 3.7659,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185074585796436e-05,
+      "loss": 3.7719,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176688638285916e-05,
+      "loss": 3.7514,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168319069579125e-05,
+      "loss": 3.7696,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159933122068605e-05,
+      "loss": 3.7764,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151547174558085e-05,
+      "loss": 3.7645,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31431776058513e-05,
+      "loss": 3.7776,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313479165834078e-05,
+      "loss": 3.7759,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312640571083026e-05,
+      "loss": 3.7886,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311801976331974e-05,
+      "loss": 3.7661,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310965019461295e-05,
+      "loss": 3.7731,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310126424710243e-05,
+      "loss": 3.7776,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309287829959191e-05,
+      "loss": 3.7623,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308449235208139e-05,
+      "loss": 3.7776,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30761227833746e-05,
+      "loss": 3.7725,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306773683586408e-05,
+      "loss": 3.7697,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305935088835356e-05,
+      "loss": 3.7769,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305096494084304e-05,
+      "loss": 3.7613,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042595372136254e-05,
+      "loss": 3.7598,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034209424625734e-05,
+      "loss": 3.7689,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025823477115214e-05,
+      "loss": 3.7732,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301743752960469e-05,
+      "loss": 3.762,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3009067960897903e-05,
+      "loss": 3.7694,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000682013387383e-05,
+      "loss": 3.7575,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299229606587686e-05,
+      "loss": 3.7651,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983910118366337e-05,
+      "loss": 3.758,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297554054965955e-05,
+      "loss": 3.7661,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296715460214903e-05,
+      "loss": 3.7791,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295878503344224e-05,
+      "loss": 3.7715,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295039908593173e-05,
+      "loss": 3.7707,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294201313842121e-05,
+      "loss": 3.7578,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293362719091069e-05,
+      "loss": 3.7622,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292524124340016e-05,
+      "loss": 3.7657,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291685529588964e-05,
+      "loss": 3.7743,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290846934837912e-05,
+      "loss": 3.7679,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29000834008686e-05,
+      "loss": 3.7709,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289171383216181e-05,
+      "loss": 3.7582,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288332788465129e-05,
+      "loss": 3.7739,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287494193714077e-05,
+      "loss": 3.758,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286655598963025e-05,
+      "loss": 3.7686,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2858186420923466e-05,
+      "loss": 3.7549,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849800473412946e-05,
+      "loss": 3.77,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284143090470616e-05,
+      "loss": 3.7557,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2833044957195635e-05,
+      "loss": 3.7633,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2824659009685115e-05,
+      "loss": 3.76,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2816273062174595e-05,
+      "loss": 3.763,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807887114664075e-05,
+      "loss": 3.7625,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799501167153555e-05,
+      "loss": 3.7597,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791115219643035e-05,
+      "loss": 3.7675,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782729272132515e-05,
+      "loss": 3.7753,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2774359703425724e-05,
+      "loss": 3.7668,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765973755915204e-05,
+      "loss": 3.768,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757587808404684e-05,
+      "loss": 3.7534,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749201860894164e-05,
+      "loss": 3.7575,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274083229218738e-05,
+      "loss": 3.7604,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273244634467686e-05,
+      "loss": 3.7682,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272406039716634e-05,
+      "loss": 3.7438,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271567444965582e-05,
+      "loss": 3.7603,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270730488094903e-05,
+      "loss": 3.7612,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269891893343851e-05,
+      "loss": 3.7608,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269053298592799e-05,
+      "loss": 3.7589,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268214703841747e-05,
+      "loss": 3.7628,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267377746971068e-05,
+      "loss": 3.7553,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266539152220016e-05,
+      "loss": 3.7678,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265700557468964e-05,
+      "loss": 3.766,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264861962717912e-05,
+      "loss": 3.7699,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264025005847233e-05,
+      "loss": 3.747,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263186411096181e-05,
+      "loss": 3.7634,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262347816345129e-05,
+      "loss": 3.7677,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261509221594077e-05,
+      "loss": 3.7556,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260672264723398e-05,
+      "loss": 3.7597,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259833669972346e-05,
+      "loss": 3.7644,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258995075221294e-05,
+      "loss": 3.7632,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258156480470242e-05,
+      "loss": 3.7768,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257321161479936e-05,
+      "loss": 3.7559,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256482566728884e-05,
+      "loss": 3.7617,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255643971977833e-05,
+      "loss": 3.7572,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254805377226781e-05,
+      "loss": 3.7574,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253966782475729e-05,
+      "loss": 3.7565,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253128187724677e-05,
+      "loss": 3.7503,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252289592973625e-05,
+      "loss": 3.7603,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251450998222573e-05,
+      "loss": 3.7416,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2506140413518936e-05,
+      "loss": 3.7641,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.875673294067383,
+      "eval_runtime": 302.8212,
+      "eval_samples_per_second": 1260.12,
+      "eval_steps_per_second": 39.38,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497754466008416e-05,
+      "loss": 3.7564,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2489368518497896e-05,
+      "loss": 3.7459,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480982570987376e-05,
+      "loss": 3.7598,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472596623476856e-05,
+      "loss": 3.7583,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464210675966336e-05,
+      "loss": 3.7684,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455824728455816e-05,
+      "loss": 3.753,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447438780945296e-05,
+      "loss": 3.7561,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2439052833434776e-05,
+      "loss": 3.7409,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243068326472799e-05,
+      "loss": 3.7539,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242229731721747e-05,
+      "loss": 3.7543,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413911369706945e-05,
+      "loss": 3.7548,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405525422196424e-05,
+      "loss": 3.748,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239717223229337e-05,
+      "loss": 3.7464,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238878628478285e-05,
+      "loss": 3.7503,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238040033727233e-05,
+      "loss": 3.7481,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237201438976181e-05,
+      "loss": 3.7377,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236362844225129e-05,
+      "loss": 3.7436,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235524249474076e-05,
+      "loss": 3.7496,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234685654723025e-05,
+      "loss": 3.7464,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233847059971973e-05,
+      "loss": 3.7612,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2330101031012945e-05,
+      "loss": 3.7548,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232171508350242e-05,
+      "loss": 3.76,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23133291359919e-05,
+      "loss": 3.7413,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230494318848138e-05,
+      "loss": 3.7489,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229657361977459e-05,
+      "loss": 3.7455,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228818767226407e-05,
+      "loss": 3.7569,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227980172475355e-05,
+      "loss": 3.7458,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227141577724303e-05,
+      "loss": 3.7476,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2263046208536236e-05,
+      "loss": 3.7409,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254660261025716e-05,
+      "loss": 3.7414,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246274313515196e-05,
+      "loss": 3.7475,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223788836600468e-05,
+      "loss": 3.7572,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222951879729789e-05,
+      "loss": 3.7532,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222113284978737e-05,
+      "loss": 3.7485,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221274690227685e-05,
+      "loss": 3.7445,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220436095476633e-05,
+      "loss": 3.7491,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219599138605954e-05,
+      "loss": 3.7511,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218760543854902e-05,
+      "loss": 3.7385,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21792194910385e-05,
+      "loss": 3.7311,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217083354352798e-05,
+      "loss": 3.7314,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216246397482119e-05,
+      "loss": 3.7401,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215407802731067e-05,
+      "loss": 3.7456,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214569207980015e-05,
+      "loss": 3.7445,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2137306132289637e-05,
+      "loss": 3.7404,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128936563582846e-05,
+      "loss": 3.741,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120550616072326e-05,
+      "loss": 3.7419,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2112164668561806e-05,
+      "loss": 3.7441,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103795099855015e-05,
+      "loss": 3.7319,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2095409152344495e-05,
+      "loss": 3.7166,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2087023204833975e-05,
+      "loss": 3.759,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078637257323454e-05,
+      "loss": 3.7421,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070267688616664e-05,
+      "loss": 3.7498,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061881741106144e-05,
+      "loss": 3.7402,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205351217239936e-05,
+      "loss": 3.739,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204512622488884e-05,
+      "loss": 3.7283,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203674027737832e-05,
+      "loss": 3.7374,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20283543298678e-05,
+      "loss": 3.7406,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201996838235728e-05,
+      "loss": 3.7322,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201158243484676e-05,
+      "loss": 3.7559,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200319648733624e-05,
+      "loss": 3.7332,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199481053982572e-05,
+      "loss": 3.7228,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198645734992266e-05,
+      "loss": 3.7482,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197807140241214e-05,
+      "loss": 3.7212,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196968545490162e-05,
+      "loss": 3.7215,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19612995073911e-05,
+      "loss": 3.7458,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195291355988058e-05,
+      "loss": 3.7424,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194452761237006e-05,
+      "loss": 3.7255,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1936141664859544e-05,
+      "loss": 3.7334,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927755717349024e-05,
+      "loss": 3.716,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191938614864223e-05,
+      "loss": 3.7312,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191100020113171e-05,
+      "loss": 3.7348,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190261425362119e-05,
+      "loss": 3.729,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189422830611067e-05,
+      "loss": 3.7374,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188584235860015e-05,
+      "loss": 3.7375,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187745641108963e-05,
+      "loss": 3.7449,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186907046357911e-05,
+      "loss": 3.7291,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186068451606859e-05,
+      "loss": 3.7348,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18523149473618e-05,
+      "loss": 3.7366,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184392899985128e-05,
+      "loss": 3.7245,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183554305234076e-05,
+      "loss": 3.7389,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182717348363398e-05,
+      "loss": 3.7354,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181878753612346e-05,
+      "loss": 3.7289,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181040158861294e-05,
+      "loss": 3.7427,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180201564110242e-05,
+      "loss": 3.72,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793646072395626e-05,
+      "loss": 3.72,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1785260124885106e-05,
+      "loss": 3.7324,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776874177374586e-05,
+      "loss": 3.7356,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1768488229864066e-05,
+      "loss": 3.7245,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760118661157275e-05,
+      "loss": 3.7319,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751732713646755e-05,
+      "loss": 3.7163,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1743346766136235e-05,
+      "loss": 3.7264,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734960818625715e-05,
+      "loss": 3.721,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172659124991893e-05,
+      "loss": 3.7291,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171820530240841e-05,
+      "loss": 3.7391,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170983573370162e-05,
+      "loss": 3.7319,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17014497861911e-05,
+      "loss": 3.7336,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169306383868058e-05,
+      "loss": 3.7217,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168467789117006e-05,
+      "loss": 3.7242,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167629194365954e-05,
+      "loss": 3.7242,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166790599614902e-05,
+      "loss": 3.7384,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165952004863849e-05,
+      "loss": 3.7314,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165113410112797e-05,
+      "loss": 3.7338,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164276453242119e-05,
+      "loss": 3.7195,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163437858491067e-05,
+      "loss": 3.7331,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162599263740015e-05,
+      "loss": 3.7221,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161760668988963e-05,
+      "loss": 3.7304,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609237121182845e-05,
+      "loss": 3.7173,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160085117367232e-05,
+      "loss": 3.732,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15924652261618e-05,
+      "loss": 3.7174,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158407927865128e-05,
+      "loss": 3.7276,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157569333114076e-05,
+      "loss": 3.7194,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156732376243397e-05,
+      "loss": 3.7326,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155893781492345e-05,
+      "loss": 3.7242,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155055186741293e-05,
+      "loss": 3.7212,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154216591990241e-05,
+      "loss": 3.733,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153377997239189e-05,
+      "loss": 3.737,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152539402488137e-05,
+      "loss": 3.7327,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151702445617458e-05,
+      "loss": 3.7277,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150863850866406e-05,
+      "loss": 3.7212,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150025256115354e-05,
+      "loss": 3.7202,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149186661364302e-05,
+      "loss": 3.7254,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14834806661325e-05,
+      "loss": 3.7268,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147509471862198e-05,
+      "loss": 3.7106,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146670877111146e-05,
+      "loss": 3.7251,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145833920240467e-05,
+      "loss": 3.7256,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144995325489415e-05,
+      "loss": 3.7253,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144156730738363e-05,
+      "loss": 3.7225,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143318135987311e-05,
+      "loss": 3.724,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142481179116632e-05,
+      "loss": 3.719,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141642584365581e-05,
+      "loss": 3.7323,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140803989614529e-05,
+      "loss": 3.7296,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139965394863477e-05,
+      "loss": 3.731,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1391284379927976e-05,
+      "loss": 3.7125,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382898432417456e-05,
+      "loss": 3.7298,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1374512484906936e-05,
+      "loss": 3.7273,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1366142916200145e-05,
+      "loss": 3.7204,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357756968689625e-05,
+      "loss": 3.7263,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349371021179105e-05,
+      "loss": 3.7261,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340985073668585e-05,
+      "loss": 3.7271,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332599126158065e-05,
+      "loss": 3.7427,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1324229557451274e-05,
+      "loss": 3.7189,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131584360994076e-05,
+      "loss": 3.725,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130745766243024e-05,
+      "loss": 3.7244,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129907171491972e-05,
+      "loss": 3.7239,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12906857674092e-05,
+      "loss": 3.7157,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128229981989868e-05,
+      "loss": 3.7141,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273913872388154e-05,
+      "loss": 3.7233,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1265527924877634e-05,
+      "loss": 3.7104,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125715835617085e-05,
+      "loss": 3.7273,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.864861011505127,
+      "eval_runtime": 305.1967,
+      "eval_samples_per_second": 1250.312,
+      "eval_steps_per_second": 39.073,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124877240866032e-05,
+      "loss": 3.7212,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12403864611498e-05,
+      "loss": 3.7096,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123200051363928e-05,
+      "loss": 3.7216,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12236309449325e-05,
+      "loss": 3.7232,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1215261376225714e-05,
+      "loss": 3.7317,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206875428715194e-05,
+      "loss": 3.7163,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1198489481204674e-05,
+      "loss": 3.724,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1190103533694154e-05,
+      "loss": 3.7107,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118171758618363e-05,
+      "loss": 3.7125,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117333163867311e-05,
+      "loss": 3.7239,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116494569116259e-05,
+      "loss": 3.7165,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115655974365207e-05,
+      "loss": 3.7116,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114820655374901e-05,
+      "loss": 3.7201,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113982060623849e-05,
+      "loss": 3.7098,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113143465872797e-05,
+      "loss": 3.7096,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112304871121745e-05,
+      "loss": 3.7051,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111466276370693e-05,
+      "loss": 3.7072,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110627681619641e-05,
+      "loss": 3.7169,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109789086868589e-05,
+      "loss": 3.7056,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108950492117537e-05,
+      "loss": 3.7292,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108113535246858e-05,
+      "loss": 3.7181,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107274940495806e-05,
+      "loss": 3.7255,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106436345744754e-05,
+      "loss": 3.7091,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105597750993702e-05,
+      "loss": 3.7127,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104760794123023e-05,
+      "loss": 3.717,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103922199371971e-05,
+      "loss": 3.7196,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103083604620919e-05,
+      "loss": 3.7101,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102245009869867e-05,
+      "loss": 3.7165,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1014080529991886e-05,
+      "loss": 3.703,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005694582481366e-05,
+      "loss": 3.71,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997308634970846e-05,
+      "loss": 3.713,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988922687460326e-05,
+      "loss": 3.7196,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980553118753535e-05,
+      "loss": 3.7194,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972167171243015e-05,
+      "loss": 3.717,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963781223732495e-05,
+      "loss": 3.7124,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955395276221975e-05,
+      "loss": 3.7142,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0947025707515184e-05,
+      "loss": 3.7186,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938639760004664e-05,
+      "loss": 3.704,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0930253812494144e-05,
+      "loss": 3.7001,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092188424378736e-05,
+      "loss": 3.6971,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091349829627684e-05,
+      "loss": 3.7053,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090511234876632e-05,
+      "loss": 3.7052,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08967264012558e-05,
+      "loss": 3.7123,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088835683254901e-05,
+      "loss": 3.7084,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087997088503849e-05,
+      "loss": 3.7038,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087158493752797e-05,
+      "loss": 3.7106,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086319899001745e-05,
+      "loss": 3.7102,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085482942131066e-05,
+      "loss": 3.6996,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084644347380014e-05,
+      "loss": 3.6847,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083805752628962e-05,
+      "loss": 3.7194,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08296715787791e-05,
+      "loss": 3.7116,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082130201007231e-05,
+      "loss": 3.7169,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081291606256179e-05,
+      "loss": 3.705,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0804546493855e-05,
+      "loss": 3.7061,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079616054634448e-05,
+      "loss": 3.6959,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078777459883396e-05,
+      "loss": 3.7011,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077938865132344e-05,
+      "loss": 3.7111,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077100270381292e-05,
+      "loss": 3.6939,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07626167563024e-05,
+      "loss": 3.7247,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075423080879188e-05,
+      "loss": 3.7055,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074584486128136e-05,
+      "loss": 3.6853,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07374916713783e-05,
+      "loss": 3.7173,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072910572386778e-05,
+      "loss": 3.6903,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072071977635726e-05,
+      "loss": 3.6883,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071233382884675e-05,
+      "loss": 3.7145,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070394788133623e-05,
+      "loss": 3.7054,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069556193382571e-05,
+      "loss": 3.6949,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068717598631519e-05,
+      "loss": 3.7006,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678790038804667e-05,
+      "loss": 3.686,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0670420470097876e-05,
+      "loss": 3.6972,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0662034522587356e-05,
+      "loss": 3.7041,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653648575076836e-05,
+      "loss": 3.6959,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0645262627566316e-05,
+      "loss": 3.7055,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636893058859525e-05,
+      "loss": 3.7017,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0628507111349005e-05,
+      "loss": 3.7132,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0620121163838485e-05,
+      "loss": 3.6974,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611735216327965e-05,
+      "loss": 3.7036,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060336564762118e-05,
+      "loss": 3.702,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059497970011066e-05,
+      "loss": 3.6935,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058659375260014e-05,
+      "loss": 3.7088,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057820780508962e-05,
+      "loss": 3.7042,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056985461518656e-05,
+      "loss": 3.6926,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056146866767604e-05,
+      "loss": 3.7101,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055308272016552e-05,
+      "loss": 3.6866,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0544696772655e-05,
+      "loss": 3.6868,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053631082514448e-05,
+      "loss": 3.7024,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052792487763396e-05,
+      "loss": 3.7002,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051953893012344e-05,
+      "loss": 3.6936,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051115298261292e-05,
+      "loss": 3.7017,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0502783413906134e-05,
+      "loss": 3.6851,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0494397466395614e-05,
+      "loss": 3.6909,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048602789768882e-05,
+      "loss": 3.6921,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04776419501783e-05,
+      "loss": 3.6953,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046925600266778e-05,
+      "loss": 3.7095,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046087005515726e-05,
+      "loss": 3.6996,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045248410764674e-05,
+      "loss": 3.7058,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044409816013622e-05,
+      "loss": 3.6887,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04357122126257e-05,
+      "loss": 3.6929,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427326265115176e-05,
+      "loss": 3.6941,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041895669640839e-05,
+      "loss": 3.7035,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041057074889787e-05,
+      "loss": 3.6977,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040218480138735e-05,
+      "loss": 3.7041,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039379885387683e-05,
+      "loss": 3.6906,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038542928517005e-05,
+      "loss": 3.6964,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037704333765953e-05,
+      "loss": 3.6954,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0368657390149e-05,
+      "loss": 3.6975,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036028782144222e-05,
+      "loss": 3.6847,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03519018739317e-05,
+      "loss": 3.7031,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0343515926421177e-05,
+      "loss": 3.6873,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033512997891065e-05,
+      "loss": 3.6939,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032674403140013e-05,
+      "loss": 3.6868,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0318374462693346e-05,
+      "loss": 3.6987,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309988515182826e-05,
+      "loss": 3.6952,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0301602567672305e-05,
+      "loss": 3.69,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0293216620161785e-05,
+      "loss": 3.6983,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284830672651265e-05,
+      "loss": 3.7071,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0276444725140745e-05,
+      "loss": 3.706,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0268075156433954e-05,
+      "loss": 3.6951,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259689208923434e-05,
+      "loss": 3.6903,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0251303261412914e-05,
+      "loss": 3.6896,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242917313902394e-05,
+      "loss": 3.6948,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0234531366391874e-05,
+      "loss": 3.6954,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0226145418881354e-05,
+      "loss": 3.6801,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217759471370834e-05,
+      "loss": 3.6957,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020938990266404e-05,
+      "loss": 3.6978,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020100395515353e-05,
+      "loss": 3.6896,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019261800764301e-05,
+      "loss": 3.6908,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018423206013249e-05,
+      "loss": 3.696,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01758624914257e-05,
+      "loss": 3.6895,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016747654391518e-05,
+      "loss": 3.7,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015909059640466e-05,
+      "loss": 3.6977,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015070464889414e-05,
+      "loss": 3.7023,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014233508018735e-05,
+      "loss": 3.6821,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013394913267683e-05,
+      "loss": 3.6973,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012556318516631e-05,
+      "loss": 3.695,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011719361645952e-05,
+      "loss": 3.6921,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108807668949e-05,
+      "loss": 3.693,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0100421721438484e-05,
+      "loss": 3.6928,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0092035773927964e-05,
+      "loss": 3.6981,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0083649826417444e-05,
+      "loss": 3.7154,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007528025771065e-05,
+      "loss": 3.6835,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006689431020013e-05,
+      "loss": 3.6966,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005850836268961e-05,
+      "loss": 3.6938,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005012241517909e-05,
+      "loss": 3.6922,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004173646766857e-05,
+      "loss": 3.6866,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003335052015805e-05,
+      "loss": 3.6835,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002496457264753e-05,
+      "loss": 3.695,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016578625137006e-05,
+      "loss": 3.6791,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000820905643022e-05,
+      "loss": 3.6967,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8586690425872803,
+      "eval_runtime": 304.6995,
+      "eval_samples_per_second": 1252.352,
+      "eval_steps_per_second": 39.137,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99998231089197e-05,
+      "loss": 3.6897,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999143716140918e-05,
+      "loss": 3.6831,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998305121389866e-05,
+      "loss": 3.6914,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997466526638814e-05,
+      "loss": 3.6925,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996629569768136e-05,
+      "loss": 3.6987,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995790975017084e-05,
+      "loss": 3.6865,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994952380266031e-05,
+      "loss": 3.6954,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994113785514979e-05,
+      "loss": 3.6779,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993275190763927e-05,
+      "loss": 3.6808,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992436596012875e-05,
+      "loss": 3.6958,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991598001261823e-05,
+      "loss": 3.6869,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990759406510771e-05,
+      "loss": 3.6794,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9899240875204655e-05,
+      "loss": 3.6917,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890854927694135e-05,
+      "loss": 3.677,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882468980183615e-05,
+      "loss": 3.6807,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9874083032673095e-05,
+      "loss": 3.6697,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865697085162575e-05,
+      "loss": 3.6801,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857311137652055e-05,
+      "loss": 3.6896,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848925190141535e-05,
+      "loss": 3.6744,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9840539242631015e-05,
+      "loss": 3.6972,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9832169673924224e-05,
+      "loss": 3.6911,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823783726413704e-05,
+      "loss": 3.6917,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9815397778903184e-05,
+      "loss": 3.6833,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9807011831392664e-05,
+      "loss": 3.6843,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979864226268587e-05,
+      "loss": 3.6837,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979025631517535e-05,
+      "loss": 3.6891,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978187036766484e-05,
+      "loss": 3.6803,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977348442015432e-05,
+      "loss": 3.6882,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976511485144753e-05,
+      "loss": 3.6745,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975672890393701e-05,
+      "loss": 3.6803,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974834295642649e-05,
+      "loss": 3.6837,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973995700891597e-05,
+      "loss": 3.6878,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973158744020918e-05,
+      "loss": 3.6925,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972320149269866e-05,
+      "loss": 3.6923,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971481554518814e-05,
+      "loss": 3.676,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970642959767762e-05,
+      "loss": 3.6858,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9698060028970827e-05,
+      "loss": 3.6899,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689674081460306e-05,
+      "loss": 3.6733,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968128813394979e-05,
+      "loss": 3.6748,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672918565243e-05,
+      "loss": 3.6698,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966453261773248e-05,
+      "loss": 3.6718,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965614667022196e-05,
+      "loss": 3.6783,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964776072271144e-05,
+      "loss": 3.6848,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963939115400465e-05,
+      "loss": 3.6803,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963100520649413e-05,
+      "loss": 3.6764,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962261925898361e-05,
+      "loss": 3.6767,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961423331147309e-05,
+      "loss": 3.6841,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96058637427663e-05,
+      "loss": 3.669,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959747779525578e-05,
+      "loss": 3.659,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958909184774526e-05,
+      "loss": 3.6883,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958070590023475e-05,
+      "loss": 3.6814,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9572336331527956e-05,
+      "loss": 3.6882,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9563950384017436e-05,
+      "loss": 3.6763,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9555580815310645e-05,
+      "loss": 3.6765,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9547194867800125e-05,
+      "loss": 3.6709,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9538808920289605e-05,
+      "loss": 3.6734,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9530422972779085e-05,
+      "loss": 3.683,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9522037025268565e-05,
+      "loss": 3.6633,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9513651077758045e-05,
+      "loss": 3.6978,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9505265130247525e-05,
+      "loss": 3.674,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496879182737005e-05,
+      "loss": 3.6553,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948852599283395e-05,
+      "loss": 3.6887,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948014004532343e-05,
+      "loss": 3.6686,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947175409781291e-05,
+      "loss": 3.6567,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946336815030239e-05,
+      "loss": 3.6821,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94549985815956e-05,
+      "loss": 3.6795,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944661263408508e-05,
+      "loss": 3.6674,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943822668657456e-05,
+      "loss": 3.6697,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942984073906404e-05,
+      "loss": 3.6561,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942145479155352e-05,
+      "loss": 3.6672,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9413068844043e-05,
+      "loss": 3.6777,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940468289653248e-05,
+      "loss": 3.6677,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939629694902196e-05,
+      "loss": 3.6757,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938792738031517e-05,
+      "loss": 3.6714,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937954143280465e-05,
+      "loss": 3.686,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937117186409786e-05,
+      "loss": 3.6713,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936278591658734e-05,
+      "loss": 3.6755,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935439996907682e-05,
+      "loss": 3.671,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93460140215663e-05,
+      "loss": 3.666,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933762807405578e-05,
+      "loss": 3.681,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932924212654526e-05,
+      "loss": 3.6755,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932087255783847e-05,
+      "loss": 3.6688,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931248661032795e-05,
+      "loss": 3.6816,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930410066281743e-05,
+      "loss": 3.6575,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929571471530691e-05,
+      "loss": 3.6632,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287328767796385e-05,
+      "loss": 3.6706,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927894282028587e-05,
+      "loss": 3.6743,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927055687277535e-05,
+      "loss": 3.6615,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926217092526483e-05,
+      "loss": 3.675,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925380135655804e-05,
+      "loss": 3.6622,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924541540904752e-05,
+      "loss": 3.6643,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923704584034074e-05,
+      "loss": 3.661,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922865989283021e-05,
+      "loss": 3.668,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922027394531969e-05,
+      "loss": 3.6816,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921188799780917e-05,
+      "loss": 3.6719,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920350205029865e-05,
+      "loss": 3.6773,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919511610278813e-05,
+      "loss": 3.6631,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918673015527761e-05,
+      "loss": 3.6651,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917834420776709e-05,
+      "loss": 3.6667,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169991017864035e-05,
+      "loss": 3.6783,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9161605070353515e-05,
+      "loss": 3.6719,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9153219122842995e-05,
+      "loss": 3.6741,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144833175332475e-05,
+      "loss": 3.6703,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9136447227821955e-05,
+      "loss": 3.6692,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9128061280311434e-05,
+      "loss": 3.6685,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9119675332800914e-05,
+      "loss": 3.6688,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9111305764094124e-05,
+      "loss": 3.6572,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102919816583604e-05,
+      "loss": 3.6729,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9094533869073083e-05,
+      "loss": 3.66,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9086147921562563e-05,
+      "loss": 3.6661,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907776197405204e-05,
+      "loss": 3.6633,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906939240534526e-05,
+      "loss": 3.6644,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906100645783474e-05,
+      "loss": 3.6701,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905262051032422e-05,
+      "loss": 3.6629,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90442345628137e-05,
+      "loss": 3.6753,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903584861530318e-05,
+      "loss": 3.6795,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902746266779266e-05,
+      "loss": 3.6771,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901907672028214e-05,
+      "loss": 3.6655,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901069077277162e-05,
+      "loss": 3.6666,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900233758286856e-05,
+      "loss": 3.6585,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899395163535804e-05,
+      "loss": 3.6683,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898556568784752e-05,
+      "loss": 3.6695,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8977179740337e-05,
+      "loss": 3.6542,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896879379282648e-05,
+      "loss": 3.6661,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8960407845315964e-05,
+      "loss": 3.6691,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8952021897805444e-05,
+      "loss": 3.667,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8943635950294924e-05,
+      "loss": 3.6617,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893528276039186e-05,
+      "loss": 3.6679,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892689681288134e-05,
+      "loss": 3.661,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891851086537082e-05,
+      "loss": 3.6756,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89101249178603e-05,
+      "loss": 3.6664,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890173897034978e-05,
+      "loss": 3.6792,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889335302283926e-05,
+      "loss": 3.6562,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888496707532874e-05,
+      "loss": 3.667,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887658112781822e-05,
+      "loss": 3.6721,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8868227937915167e-05,
+      "loss": 3.666,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8859841990404647e-05,
+      "loss": 3.6647,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8851456042894126e-05,
+      "loss": 3.6681,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8843070095383606e-05,
+      "loss": 3.6698,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8834684147873086e-05,
+      "loss": 3.6851,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8826314579166295e-05,
+      "loss": 3.6627,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817928631655775e-05,
+      "loss": 3.669,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8809542684145255e-05,
+      "loss": 3.6682,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8801156736634735e-05,
+      "loss": 3.6638,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792770789124215e-05,
+      "loss": 3.6609,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8784401220417424e-05,
+      "loss": 3.657,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8776015272906904e-05,
+      "loss": 3.665,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8767629325396384e-05,
+      "loss": 3.6587,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875924337788587e-05,
+      "loss": 3.6658,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8542556762695312,
+      "eval_runtime": 308.0052,
+      "eval_samples_per_second": 1238.911,
+      "eval_steps_per_second": 38.717,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750857430375344e-05,
+      "loss": 3.6634,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8742471482864824e-05,
+      "loss": 3.6554,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8734085535354304e-05,
+      "loss": 3.6598,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872571596664752e-05,
+      "loss": 3.6708,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871733001913699e-05,
+      "loss": 3.6766,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870894407162647e-05,
+      "loss": 3.6608,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870055812411595e-05,
+      "loss": 3.6635,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869218855540916e-05,
+      "loss": 3.6548,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868380260789864e-05,
+      "loss": 3.6549,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867541666038812e-05,
+      "loss": 3.6657,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866703071287761e-05,
+      "loss": 3.6635,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865864476536709e-05,
+      "loss": 3.6573,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86502751966603e-05,
+      "loss": 3.6652,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864188924914978e-05,
+      "loss": 3.6507,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863350330163926e-05,
+      "loss": 3.6536,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862511735412874e-05,
+      "loss": 3.6421,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861673140661822e-05,
+      "loss": 3.6586,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86083454591077e-05,
+      "loss": 3.6613,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859995951159718e-05,
+      "loss": 3.6488,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859157356408666e-05,
+      "loss": 3.67,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858320399537987e-05,
+      "loss": 3.6644,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857481804786935e-05,
+      "loss": 3.6686,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8566432100358827e-05,
+      "loss": 3.6581,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8558046152848307e-05,
+      "loss": 3.6563,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854967658414152e-05,
+      "loss": 3.6579,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8541290636631e-05,
+      "loss": 3.6667,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853290468912048e-05,
+      "loss": 3.6485,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852451874160996e-05,
+      "loss": 3.6639,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851614917290317e-05,
+      "loss": 3.6478,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850776322539265e-05,
+      "loss": 3.6536,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849937727788213e-05,
+      "loss": 3.6602,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849099133037161e-05,
+      "loss": 3.664,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848262176166482e-05,
+      "loss": 3.6577,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84742358141543e-05,
+      "loss": 3.6698,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846584986664378e-05,
+      "loss": 3.6535,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845746391913326e-05,
+      "loss": 3.6582,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8449094350426476e-05,
+      "loss": 3.6634,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8440708402915956e-05,
+      "loss": 3.6456,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8432322455405436e-05,
+      "loss": 3.648,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423936507894916e-05,
+      "loss": 3.6479,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8415566939188125e-05,
+      "loss": 3.6467,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8407180991677605e-05,
+      "loss": 3.65,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8398795044167085e-05,
+      "loss": 3.6636,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8390409096656565e-05,
+      "loss": 3.6525,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8382039527949774e-05,
+      "loss": 3.6518,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837366995924298e-05,
+      "loss": 3.6511,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836528401173246e-05,
+      "loss": 3.657,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835689806422195e-05,
+      "loss": 3.6422,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834851211671143e-05,
+      "loss": 3.6345,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834012616920091e-05,
+      "loss": 3.6637,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833174022169039e-05,
+      "loss": 3.657,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832335427417987e-05,
+      "loss": 3.6653,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831496832666935e-05,
+      "loss": 3.6486,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830661513676629e-05,
+      "loss": 3.6506,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829822918925577e-05,
+      "loss": 3.6448,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828984324174525e-05,
+      "loss": 3.6461,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828145729423473e-05,
+      "loss": 3.6613,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827307134672421e-05,
+      "loss": 3.6338,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826468539921369e-05,
+      "loss": 3.6751,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825629945170317e-05,
+      "loss": 3.6531,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824791350419265e-05,
+      "loss": 3.6277,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823956031428959e-05,
+      "loss": 3.6586,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823117436677907e-05,
+      "loss": 3.645,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822278841926855e-05,
+      "loss": 3.6328,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821440247175803e-05,
+      "loss": 3.6541,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820603290305124e-05,
+      "loss": 3.6581,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819764695554072e-05,
+      "loss": 3.6399,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81892610080302e-05,
+      "loss": 3.6473,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818087506051968e-05,
+      "loss": 3.6298,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817248911300916e-05,
+      "loss": 3.6436,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816410316549864e-05,
+      "loss": 3.6509,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815573359679186e-05,
+      "loss": 3.6416,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814734764928134e-05,
+      "loss": 3.6533,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813896170177082e-05,
+      "loss": 3.6439,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81305757542603e-05,
+      "loss": 3.6637,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812218980674977e-05,
+      "loss": 3.6441,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811380385923925e-05,
+      "loss": 3.6537,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810541791172873e-05,
+      "loss": 3.647,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809703196421821e-05,
+      "loss": 3.6408,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808866239551142e-05,
+      "loss": 3.654,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80802764480009e-05,
+      "loss": 3.6469,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071906879294115e-05,
+      "loss": 3.6513,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8063520931783595e-05,
+      "loss": 3.6521,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8055134984273075e-05,
+      "loss": 3.6398,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8046749036762555e-05,
+      "loss": 3.6302,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8038363089252035e-05,
+      "loss": 3.6455,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029993520545244e-05,
+      "loss": 3.6527,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8021607573034724e-05,
+      "loss": 3.6385,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8013221625524204e-05,
+      "loss": 3.649,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8004835678013684e-05,
+      "loss": 3.6368,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7996449730503164e-05,
+      "loss": 3.6402,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798808016179637e-05,
+      "loss": 3.6389,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797969421428585e-05,
+      "loss": 3.6429,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797130826677533e-05,
+      "loss": 3.6562,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796292231926481e-05,
+      "loss": 3.6484,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795453637175429e-05,
+      "loss": 3.6532,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794616680304751e-05,
+      "loss": 3.6378,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793778085553699e-05,
+      "loss": 3.6404,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792939490802647e-05,
+      "loss": 3.6427,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792100896051595e-05,
+      "loss": 3.6515,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791263939180916e-05,
+      "loss": 3.6465,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790425344429864e-05,
+      "loss": 3.6507,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789586749678812e-05,
+      "loss": 3.6453,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78874815492776e-05,
+      "loss": 3.6414,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787909560176708e-05,
+      "loss": 3.6467,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787070965425656e-05,
+      "loss": 3.6429,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7862340085549766e-05,
+      "loss": 3.6377,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7853954138039246e-05,
+      "loss": 3.6465,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7845568190528726e-05,
+      "loss": 3.6377,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783718224301821e-05,
+      "loss": 3.6407,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782879629550769e-05,
+      "loss": 3.6389,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78204267268009e-05,
+      "loss": 3.6395,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781204077929038e-05,
+      "loss": 3.6495,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780365483177986e-05,
+      "loss": 3.6394,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779526888426934e-05,
+      "loss": 3.6473,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778688293675882e-05,
+      "loss": 3.6519,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77784969892483e-05,
+      "loss": 3.6532,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777012742054151e-05,
+      "loss": 3.6433,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776174147303099e-05,
+      "loss": 3.6436,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775335552552047e-05,
+      "loss": 3.6347,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774496957800995e-05,
+      "loss": 3.6481,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773658363049943e-05,
+      "loss": 3.6436,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772819768298891e-05,
+      "loss": 3.6285,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771981173547839e-05,
+      "loss": 3.6412,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7711442166771607e-05,
+      "loss": 3.6429,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770305621926108e-05,
+      "loss": 3.6461,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7694686650554296e-05,
+      "loss": 3.6401,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7686300703043776e-05,
+      "loss": 3.6422,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7677914755533255e-05,
+      "loss": 3.636,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766952880802273e-05,
+      "loss": 3.6515,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766114286051221e-05,
+      "loss": 3.6438,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7652773291805424e-05,
+      "loss": 3.655,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7644387344294904e-05,
+      "loss": 3.6332,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7636001396784384e-05,
+      "loss": 3.6435,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7627615449273864e-05,
+      "loss": 3.6487,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761924588056708e-05,
+      "loss": 3.6378,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7610859933056553e-05,
+      "loss": 3.643,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760247398554603e-05,
+      "loss": 3.642,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759408803803551e-05,
+      "loss": 3.6496,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758570209052499e-05,
+      "loss": 3.6552,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75773325218182e-05,
+      "loss": 3.6437,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756894657430768e-05,
+      "loss": 3.6422,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756056062679716e-05,
+      "loss": 3.6459,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755217467928664e-05,
+      "loss": 3.6412,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754378873177612e-05,
+      "loss": 3.6378,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753541916306934e-05,
+      "loss": 3.6345,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752703321555882e-05,
+      "loss": 3.6406,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75186472680483e-05,
+      "loss": 3.6352,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751026132053778e-05,
+      "loss": 3.6407,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8511030673980713,
+      "eval_runtime": 320.7121,
+      "eval_samples_per_second": 1189.824,
+      "eval_steps_per_second": 37.183,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750187537302726e-05,
+      "loss": 3.6437,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749348942551674e-05,
+      "loss": 3.6315,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748510347800622e-05,
+      "loss": 3.6363,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74767175304957e-05,
+      "loss": 3.6453,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746834796178891e-05,
+      "loss": 3.6526,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745996201427839e-05,
+      "loss": 3.6388,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745157606676787e-05,
+      "loss": 3.645,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744319011925735e-05,
+      "loss": 3.6325,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743480417174683e-05,
+      "loss": 3.6302,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742641822423631e-05,
+      "loss": 3.6386,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7418032276725787e-05,
+      "loss": 3.6434,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7409646329215267e-05,
+      "loss": 3.6347,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740129313931221e-05,
+      "loss": 3.6404,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739290719180169e-05,
+      "loss": 3.6253,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738452124429117e-05,
+      "loss": 3.6313,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737613529678065e-05,
+      "loss": 3.6165,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736774934927013e-05,
+      "loss": 3.6369,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735936340175961e-05,
+      "loss": 3.6365,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735099383305282e-05,
+      "loss": 3.6262,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73426078855423e-05,
+      "loss": 3.6447,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733422193803178e-05,
+      "loss": 3.644,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732583599052126e-05,
+      "loss": 3.6445,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731745004301074e-05,
+      "loss": 3.6338,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730906409550022e-05,
+      "loss": 3.6342,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73006781479897e-05,
+      "loss": 3.6381,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729229220047918e-05,
+      "loss": 3.642,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728392263177239e-05,
+      "loss": 3.6272,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727553668426187e-05,
+      "loss": 3.64,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7267167115555085e-05,
+      "loss": 3.6295,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7258781168044565e-05,
+      "loss": 3.6267,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.725039522053404e-05,
+      "loss": 3.6353,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724200927302352e-05,
+      "loss": 3.6416,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7233623325513e-05,
+      "loss": 3.6367,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722523737800248e-05,
+      "loss": 3.645,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721685143049196e-05,
+      "loss": 3.6296,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7208465482981445e-05,
+      "loss": 3.6376,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7200095914274654e-05,
+      "loss": 3.6433,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7191709966764134e-05,
+      "loss": 3.6223,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7183324019253614e-05,
+      "loss": 3.6242,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7174938071743094e-05,
+      "loss": 3.6263,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71665685030363e-05,
+      "loss": 3.6199,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715818255552578e-05,
+      "loss": 3.6294,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714979660801526e-05,
+      "loss": 3.6423,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714141066050474e-05,
+      "loss": 3.6275,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713305747060168e-05,
+      "loss": 3.6306,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712467152309117e-05,
+      "loss": 3.6316,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711628557558065e-05,
+      "loss": 3.6343,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710789962807013e-05,
+      "loss": 3.6207,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709951368055961e-05,
+      "loss": 3.6117,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709112773304909e-05,
+      "loss": 3.638,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708274178553857e-05,
+      "loss": 3.6357,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707435583802805e-05,
+      "loss": 3.6444,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7065986269321256e-05,
+      "loss": 3.6277,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7057616700614466e-05,
+      "loss": 3.6267,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7049230753103946e-05,
+      "loss": 3.6235,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7040844805593425e-05,
+      "loss": 3.6236,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7032458858082905e-05,
+      "loss": 3.6374,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7024072910572385e-05,
+      "loss": 3.6095,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7015686963061865e-05,
+      "loss": 3.6516,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700730101555135e-05,
+      "loss": 3.6321,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699891506804083e-05,
+      "loss": 3.6086,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699056187813777e-05,
+      "loss": 3.6345,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698217593062725e-05,
+      "loss": 3.6222,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697378998311673e-05,
+      "loss": 3.6104,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696540403560621e-05,
+      "loss": 3.6324,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695703446689942e-05,
+      "loss": 3.6347,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69486485193889e-05,
+      "loss": 3.6154,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694026257187838e-05,
+      "loss": 3.63,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693187662436786e-05,
+      "loss": 3.6051,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692349067685734e-05,
+      "loss": 3.6215,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691510472934682e-05,
+      "loss": 3.6234,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6906735160640035e-05,
+      "loss": 3.6215,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6898349213129515e-05,
+      "loss": 3.6274,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6889963265618995e-05,
+      "loss": 3.6231,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6881577318108475e-05,
+      "loss": 3.6426,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6873207749401684e-05,
+      "loss": 3.6243,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6864821801891164e-05,
+      "loss": 3.6326,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6856435854380644e-05,
+      "loss": 3.6215,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6848049906870124e-05,
+      "loss": 3.6197,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6839663959359604e-05,
+      "loss": 3.6348,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6831278011849084e-05,
+      "loss": 3.6247,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682290844314229e-05,
+      "loss": 3.6259,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681452249563177e-05,
+      "loss": 3.6334,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680613654812126e-05,
+      "loss": 3.6193,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679775060061074e-05,
+      "loss": 3.6086,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678936465310022e-05,
+      "loss": 3.622,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678099508439343e-05,
+      "loss": 3.6317,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677260913688291e-05,
+      "loss": 3.6131,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676422318937239e-05,
+      "loss": 3.6287,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675583724186187e-05,
+      "loss": 3.6198,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674745129435134e-05,
+      "loss": 3.6128,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673908172564456e-05,
+      "loss": 3.6179,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673069577813404e-05,
+      "loss": 3.6207,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672230983062352e-05,
+      "loss": 3.6385,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6713923883113e-05,
+      "loss": 3.6247,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670553793560248e-05,
+      "loss": 3.6296,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669715198809196e-05,
+      "loss": 3.619,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668876604058144e-05,
+      "loss": 3.6164,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6680396471874646e-05,
+      "loss": 3.6192,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6672010524364126e-05,
+      "loss": 3.6301,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6663624576853606e-05,
+      "loss": 3.6281,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6655238629343086e-05,
+      "loss": 3.6274,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6646852681832566e-05,
+      "loss": 3.6241,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6638466734322046e-05,
+      "loss": 3.6166,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6630080786811526e-05,
+      "loss": 3.6287,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6621711218104735e-05,
+      "loss": 3.6215,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6613325270594215e-05,
+      "loss": 3.6167,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6604939323083695e-05,
+      "loss": 3.622,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6596553375573175e-05,
+      "loss": 3.6196,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658816742806266e-05,
+      "loss": 3.6193,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657979785935587e-05,
+      "loss": 3.6154,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657141191184535e-05,
+      "loss": 3.6164,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656302596433483e-05,
+      "loss": 3.6257,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655464001682431e-05,
+      "loss": 3.6237,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654625406931379e-05,
+      "loss": 3.6205,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653786812180327e-05,
+      "loss": 3.6299,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652948217429275e-05,
+      "loss": 3.6348,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652111260558596e-05,
+      "loss": 3.6189,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651272665807544e-05,
+      "loss": 3.6237,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650434071056492e-05,
+      "loss": 3.6128,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64959547630544e-05,
+      "loss": 3.6248,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648756881554388e-05,
+      "loss": 3.6216,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6479199246837095e-05,
+      "loss": 3.6129,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6470813299326575e-05,
+      "loss": 3.6148,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6462427351816055e-05,
+      "loss": 3.6235,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6454041404305535e-05,
+      "loss": 3.6248,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644565545679501e-05,
+      "loss": 3.6145,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643726950928449e-05,
+      "loss": 3.6196,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6428899940577704e-05,
+      "loss": 3.6188,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642051399306718e-05,
+      "loss": 3.6262,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641212804555666e-05,
+      "loss": 3.6239,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640374209804614e-05,
+      "loss": 3.6337,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639537252933935e-05,
+      "loss": 3.611,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638698658182883e-05,
+      "loss": 3.6247,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637860063431831e-05,
+      "loss": 3.6254,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637021468680779e-05,
+      "loss": 3.6142,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636182873929727e-05,
+      "loss": 3.6218,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635345917059048e-05,
+      "loss": 3.6238,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634507322307996e-05,
+      "loss": 3.6302,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633668727556944e-05,
+      "loss": 3.6329,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632830132805892e-05,
+      "loss": 3.6209,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631993175935213e-05,
+      "loss": 3.6212,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631154581184161e-05,
+      "loss": 3.6252,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630315986433109e-05,
+      "loss": 3.6186,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629477391682057e-05,
+      "loss": 3.6168,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628640434811379e-05,
+      "loss": 3.6144,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627801840060327e-05,
+      "loss": 3.6151,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6269632453092747e-05,
+      "loss": 3.6187,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6261246505582227e-05,
+      "loss": 3.614,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.849750280380249,
+      "eval_runtime": 302.9113,
+      "eval_samples_per_second": 1259.745,
+      "eval_steps_per_second": 39.368,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6252860558071707e-05,
+      "loss": 3.6213,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6244474610561186e-05,
+      "loss": 3.611,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6236088663050666e-05,
+      "loss": 3.6176,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6227719094343876e-05,
+      "loss": 3.624,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6219333146833355e-05,
+      "loss": 3.6296,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6210947199322835e-05,
+      "loss": 3.618,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6202561251812315e-05,
+      "loss": 3.6251,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6194175304301795e-05,
+      "loss": 3.6143,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6185789356791275e-05,
+      "loss": 3.6087,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6177403409280755e-05,
+      "loss": 3.6146,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6169017461770235e-05,
+      "loss": 3.6261,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616066427186718e-05,
+      "loss": 3.6096,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615229470316039e-05,
+      "loss": 3.6237,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614390875564987e-05,
+      "loss": 3.6026,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613552280813935e-05,
+      "loss": 3.6075,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612713686062883e-05,
+      "loss": 3.5979,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611875091311831e-05,
+      "loss": 3.6179,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611036496560779e-05,
+      "loss": 3.6138,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610197901809727e-05,
+      "loss": 3.6093,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609359307058675e-05,
+      "loss": 3.6151,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608520712307623e-05,
+      "loss": 3.6275,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607682117556571e-05,
+      "loss": 3.6208,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606843522805519e-05,
+      "loss": 3.6159,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606004928054467e-05,
+      "loss": 3.6116,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6051679711837885e-05,
+      "loss": 3.6213,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6043293764327365e-05,
+      "loss": 3.616,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603490781681684e-05,
+      "loss": 3.6109,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602652186930632e-05,
+      "loss": 3.6171,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601816867940326e-05,
+      "loss": 3.6068,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600978273189274e-05,
+      "loss": 3.6039,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600139678438222e-05,
+      "loss": 3.6191,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59930108368717e-05,
+      "loss": 3.6153,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598462488936118e-05,
+      "loss": 3.6214,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597623894185066e-05,
+      "loss": 3.6218,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596785299434014e-05,
+      "loss": 3.6086,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595948342563336e-05,
+      "loss": 3.6169,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595109747812284e-05,
+      "loss": 3.6204,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594271153061231e-05,
+      "loss": 3.6059,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593432558310179e-05,
+      "loss": 3.6074,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592593963559127e-05,
+      "loss": 3.5995,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591755368808075e-05,
+      "loss": 3.6021,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590916774057023e-05,
+      "loss": 3.6096,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590078179305971e-05,
+      "loss": 3.622,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589241222435292e-05,
+      "loss": 3.6085,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5884042655646136e-05,
+      "loss": 3.6064,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5875656708135616e-05,
+      "loss": 3.6119,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5867270760625096e-05,
+      "loss": 3.6121,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5858884813114576e-05,
+      "loss": 3.603,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5850498865604056e-05,
+      "loss": 3.5921,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5842129296897265e-05,
+      "loss": 3.615,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5833743349386745e-05,
+      "loss": 3.62,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5825357401876225e-05,
+      "loss": 3.6185,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5816971454365705e-05,
+      "loss": 3.6075,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5808601885658914e-05,
+      "loss": 3.604,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5800215938148394e-05,
+      "loss": 3.6018,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5791829990637874e-05,
+      "loss": 3.6009,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5783444043127354e-05,
+      "loss": 3.6176,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5775058095616834e-05,
+      "loss": 3.594,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5766672148106314e-05,
+      "loss": 3.6288,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57582862005958e-05,
+      "loss": 3.6105,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574990025308528e-05,
+      "loss": 3.5925,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574154706318222e-05,
+      "loss": 3.6103,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57331611156717e-05,
+      "loss": 3.6034,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572477516816118e-05,
+      "loss": 3.592,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571638922065066e-05,
+      "loss": 3.6075,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570801965194387e-05,
+      "loss": 3.6162,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569963370443335e-05,
+      "loss": 3.5967,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569124775692283e-05,
+      "loss": 3.6098,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568286180941231e-05,
+      "loss": 3.5863,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567447586190179e-05,
+      "loss": 3.6017,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566608991439127e-05,
+      "loss": 3.6047,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5657720345684483e-05,
+      "loss": 3.6039,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5649334398173963e-05,
+      "loss": 3.6042,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5640948450663443e-05,
+      "loss": 3.604,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563256250315292e-05,
+      "loss": 3.6188,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562419293444613e-05,
+      "loss": 3.6061,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561580698693561e-05,
+      "loss": 3.6096,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560742103942509e-05,
+      "loss": 3.6037,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559903509191457e-05,
+      "loss": 3.5992,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559064914440405e-05,
+      "loss": 3.6153,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558226319689353e-05,
+      "loss": 3.6042,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557389362818674e-05,
+      "loss": 3.6037,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556550768067622e-05,
+      "loss": 3.6146,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555712173316571e-05,
+      "loss": 3.6019,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554873578565519e-05,
+      "loss": 3.5858,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554034983814467e-05,
+      "loss": 3.6035,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553198026943788e-05,
+      "loss": 3.6079,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552359432192736e-05,
+      "loss": 3.5963,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551520837441684e-05,
+      "loss": 3.6052,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550682242690632e-05,
+      "loss": 3.6021,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54984364793958e-05,
+      "loss": 3.5928,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5490066910689006e-05,
+      "loss": 3.6003,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5481680963178486e-05,
+      "loss": 3.5977,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5473295015667966e-05,
+      "loss": 3.6142,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5464909068157446e-05,
+      "loss": 3.6077,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5456523120646926e-05,
+      "loss": 3.61,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5448137173136406e-05,
+      "loss": 3.597,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5439751225625886e-05,
+      "loss": 3.5962,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5431365278115366e-05,
+      "loss": 3.602,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5422995709408575e-05,
+      "loss": 3.6101,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5414609761898055e-05,
+      "loss": 3.6062,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5406223814387535e-05,
+      "loss": 3.6077,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5397837866877015e-05,
+      "loss": 3.6016,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5389451919366495e-05,
+      "loss": 3.5957,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5381065971855975e-05,
+      "loss": 3.6083,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5372696403149184e-05,
+      "loss": 3.6004,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5364310455638664e-05,
+      "loss": 3.6016,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5355924508128144e-05,
+      "loss": 3.5997,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5347538560617623e-05,
+      "loss": 3.5977,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533915261310711e-05,
+      "loss": 3.5984,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533076666559659e-05,
+      "loss": 3.5946,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53223970968898e-05,
+      "loss": 3.5949,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531401114937928e-05,
+      "loss": 3.6087,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530562520186876e-05,
+      "loss": 3.6041,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529723925435824e-05,
+      "loss": 3.5983,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528885330684772e-05,
+      "loss": 3.609,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52804673593372e-05,
+      "loss": 3.6184,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527209779063041e-05,
+      "loss": 3.5979,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526371184311989e-05,
+      "loss": 3.5999,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525532589560937e-05,
+      "loss": 3.5975,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524693994809885e-05,
+      "loss": 3.6042,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523855400058833e-05,
+      "loss": 3.5994,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5230184431881544e-05,
+      "loss": 3.5962,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5221798484371024e-05,
+      "loss": 3.5963,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5213412536860504e-05,
+      "loss": 3.6059,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5205026589349984e-05,
+      "loss": 3.5991,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519665702064319e-05,
+      "loss": 3.5969,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518827107313267e-05,
+      "loss": 3.6054,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517988512562215e-05,
+      "loss": 3.6008,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517149917811163e-05,
+      "loss": 3.6018,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5163113230601106e-05,
+      "loss": 3.6062,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515474366189432e-05,
+      "loss": 3.6117,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51463577143838e-05,
+      "loss": 3.5965,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513797176687328e-05,
+      "loss": 3.6041,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512958581936276e-05,
+      "loss": 3.6045,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512121625065598e-05,
+      "loss": 3.5965,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511283030314546e-05,
+      "loss": 3.5984,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510444435563493e-05,
+      "loss": 3.6037,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509605840812441e-05,
+      "loss": 3.6135,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508767246061389e-05,
+      "loss": 3.6124,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50793028919071e-05,
+      "loss": 3.603,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507091694439658e-05,
+      "loss": 3.6019,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506253099688606e-05,
+      "loss": 3.6062,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505414504937554e-05,
+      "loss": 3.6023,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504575910186502e-05,
+      "loss": 3.5946,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5037389533158235e-05,
+      "loss": 3.5996,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5029003585647715e-05,
+      "loss": 3.5912,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5020617638137195e-05,
+      "loss": 3.6,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5012231690626675e-05,
+      "loss": 3.5939,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8499250411987305,
+      "eval_runtime": 302.7141,
+      "eval_samples_per_second": 1260.566,
+      "eval_steps_per_second": 39.394,
+      "step": 915840
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.254061852335432e+17,
+  "trial_name": null,
+  "trial_params": null
+}