diff --git "a/checkpoint-457914/trainer_state.json" "b/checkpoint-457914/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457914/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 4.164071083068848,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl/lstm/4/checkpoints/checkpoint-457914",
+  "epoch": 0.025000278439663435,
+  "eval_steps": 10,
+  "global_step": 457914,
+  "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": 10.8208,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5681,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0543,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9781,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9513,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.853,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.725,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6242,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5398,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4465,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3928,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3159,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.243,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1756,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1279,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0907,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.049,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 5.995,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9568,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.9174,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8778,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8494,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8176,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.784,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7541,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7369,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7035,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6817,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6452,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6414,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6096,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5922,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5836,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5653,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5353,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5227,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 5.5151,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 5.4915,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 5.4802,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.4601,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4353,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4248,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4113,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.403,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.3914,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.386,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3616,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3616,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 5.3657,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3372,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3266,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3148,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 5.3236,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955570856997975e-05,
+      "loss": 5.2862,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 5.3039,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 5.2737,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 5.2707,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 5.2651,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 5.2549,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950540926372037e-05,
+      "loss": 5.2412,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949702331620985e-05,
+      "loss": 5.2298,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 5.2154,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 5.222,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 5.2089,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946349590497149e-05,
+      "loss": 5.1993,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 5.1913,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 5.182,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 5.1639,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 5.1817,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 5.159,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 5.1512,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 5.16,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 5.136,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 5.1494,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 5.1395,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 5.1287,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 5.1284,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 5.1021,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 5.1005,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 5.0972,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 5.0964,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932100031250758e-05,
+      "loss": 5.0856,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931261436499706e-05,
+      "loss": 5.0738,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930422841748654e-05,
+      "loss": 5.0588,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 5.0659,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 5.0642,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 5.0614,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 5.0583,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 5.0528,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 5.0403,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 5.0224,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 5.0187,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 5.0257,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 5.0202,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.9978,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 5.0022,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 5.0014,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 5.0008,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.9918,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.9929,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.9855,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.9765,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.9745,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.9757,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.9602,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9637,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.9405,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.9354,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.9348,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.9409,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.9314,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.9087,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.9141,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.919,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.9243,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9111,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.9118,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.9058,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.9083,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.8923,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.8944,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.8834,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.8886,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.8694,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.8835,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.8724,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.8802,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.8519,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.8641,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.8732,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.8464,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.8651,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8492,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8514,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8419,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.8377,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.8368,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.8365,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.8361,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.8294,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.8323,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.8305,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.8108,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8115,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.8025,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8038,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.8039,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.7967,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7955,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7938,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.755862236022949,
+      "eval_runtime": 556.7302,
+      "eval_samples_per_second": 685.415,
+      "eval_steps_per_second": 21.42,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.8042,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.795,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7841,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7864,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.779,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7572,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7721,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7738,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.7557,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.7694,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8658690626017536e-05,
+      "loss": 4.7739,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8650304678507016e-05,
+      "loss": 4.7575,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8641918730996496e-05,
+      "loss": 4.7523,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8633532783485976e-05,
+      "loss": 4.7528,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8625146835975456e-05,
+      "loss": 4.7537,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8616760888464936e-05,
+      "loss": 4.7464,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8608391319758145e-05,
+      "loss": 4.7423,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7406,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.7362,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8583249856030314e-05,
+      "loss": 4.7415,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574863908519794e-05,
+      "loss": 4.7402,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566477961009274e-05,
+      "loss": 4.7392,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558092013498754e-05,
+      "loss": 4.7343,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854970606598824e-05,
+      "loss": 4.733,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854132011847772e-05,
+      "loss": 4.7277,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85329341709672e-05,
+      "loss": 4.7248,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.852456460226041e-05,
+      "loss": 4.7267,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.851617865474989e-05,
+      "loss": 4.7086,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850779270723937e-05,
+      "loss": 4.7228,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849940675972885e-05,
+      "loss": 4.7032,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849103719102206e-05,
+      "loss": 4.7088,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.848265124351154e-05,
+      "loss": 4.7177,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847426529600102e-05,
+      "loss": 4.7162,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84658793484905e-05,
+      "loss": 4.7015,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.693,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8449123832273194e-05,
+      "loss": 4.7097,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8440737884762674e-05,
+      "loss": 4.6982,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8432351937252154e-05,
+      "loss": 4.6967,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8423965989741634e-05,
+      "loss": 4.6938,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8415580042231114e-05,
+      "loss": 4.6789,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840719409472059e-05,
+      "loss": 4.6769,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83988245260138e-05,
+      "loss": 4.6778,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839043857850328e-05,
+      "loss": 4.6822,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.838205263099276e-05,
+      "loss": 4.6756,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8373666683482236e-05,
+      "loss": 4.6777,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836529711477545e-05,
+      "loss": 4.671,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835691116726493e-05,
+      "loss": 4.6728,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834852521975441e-05,
+      "loss": 4.6835,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834013927224389e-05,
+      "loss": 4.6646,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833175332473337e-05,
+      "loss": 4.6665,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6554,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6776,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6471,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6699,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6472,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828147039727771e-05,
+      "loss": 4.6551,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.827308444976719e-05,
+      "loss": 4.655,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.826469850225667e-05,
+      "loss": 4.6509,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.825631255474615e-05,
+      "loss": 4.6423,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.824792660723563e-05,
+      "loss": 4.641,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823954065972511e-05,
+      "loss": 4.6303,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6456,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6371,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6306,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.6304,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8197643679779975e-05,
+      "loss": 4.6295,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8189257732269455e-05,
+      "loss": 4.6114,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8180871784758935e-05,
+      "loss": 4.6367,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8172485837248415e-05,
+      "loss": 4.623,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8164099889737895e-05,
+      "loss": 4.6217,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8155713942227375e-05,
+      "loss": 4.6306,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8147327994716854e-05,
+      "loss": 4.616,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.63,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813057247849955e-05,
+      "loss": 4.6321,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.812218653098903e-05,
+      "loss": 4.6234,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.811380058347851e-05,
+      "loss": 4.624,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.810541463596799e-05,
+      "loss": 4.6059,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.809702868845747e-05,
+      "loss": 4.6017,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808864274094695e-05,
+      "loss": 4.6133,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808027317224016e-05,
+      "loss": 4.6147,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.807188722472964e-05,
+      "loss": 4.6061,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.806350127721912e-05,
+      "loss": 4.6017,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.80551153297086e-05,
+      "loss": 4.5863,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804672938219807e-05,
+      "loss": 4.5953,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803834343468755e-05,
+      "loss": 4.5994,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802997386598077e-05,
+      "loss": 4.6007,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802158791847025e-05,
+      "loss": 4.6059,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801320197095973e-05,
+      "loss": 4.6049,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800481602344921e-05,
+      "loss": 4.5933,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799643007593869e-05,
+      "loss": 4.5851,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798804412842817e-05,
+      "loss": 4.5802,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797965818091765e-05,
+      "loss": 4.5868,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797127223340713e-05,
+      "loss": 4.5881,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796290266470034e-05,
+      "loss": 4.5758,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795451671718982e-05,
+      "loss": 4.5771,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.79461307696793e-05,
+      "loss": 4.5847,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793774482216878e-05,
+      "loss": 4.5891,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792935887465826e-05,
+      "loss": 4.5774,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792098930595147e-05,
+      "loss": 4.5832,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791260335844095e-05,
+      "loss": 4.5845,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.790421741093043e-05,
+      "loss": 4.5767,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789583146341991e-05,
+      "loss": 4.5713,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788744551590939e-05,
+      "loss": 4.5821,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.78790759472026e-05,
+      "loss": 4.5684,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787068999969208e-05,
+      "loss": 4.5789,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.786230405218156e-05,
+      "loss": 4.5601,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.785391810467104e-05,
+      "loss": 4.5503,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.784554853596425e-05,
+      "loss": 4.5574,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.783716258845373e-05,
+      "loss": 4.5643,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782877664094321e-05,
+      "loss": 4.558,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782039069343269e-05,
+      "loss": 4.5463,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.781200474592217e-05,
+      "loss": 4.5446,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780361879841165e-05,
+      "loss": 4.5538,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.779523285090114e-05,
+      "loss": 4.5647,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7786863282194346e-05,
+      "loss": 4.5517,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7778477334683826e-05,
+      "loss": 4.5597,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7770091387173306e-05,
+      "loss": 4.5555,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7761705439662786e-05,
+      "loss": 4.5514,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775331949215226e-05,
+      "loss": 4.5476,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774493354464174e-05,
+      "loss": 4.5453,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7736563975934955e-05,
+      "loss": 4.5405,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7728178028424435e-05,
+      "loss": 4.5509,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771979208091391e-05,
+      "loss": 4.536,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771140613340339e-05,
+      "loss": 4.5382,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7703036564696604e-05,
+      "loss": 4.5416,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7694650617186084e-05,
+      "loss": 4.5494,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7686264669675564e-05,
+      "loss": 4.5264,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7677878722165044e-05,
+      "loss": 4.5314,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7669492774654524e-05,
+      "loss": 4.5525,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7661106827144004e-05,
+      "loss": 4.526,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7652720879633484e-05,
+      "loss": 4.5463,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7644334932122964e-05,
+      "loss": 4.5298,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7635948984612444e-05,
+      "loss": 4.5414,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762757941590565e-05,
+      "loss": 4.5284,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761919346839513e-05,
+      "loss": 4.5239,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761080752088461e-05,
+      "loss": 4.53,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760242157337409e-05,
+      "loss": 4.528,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759403562586357e-05,
+      "loss": 4.5353,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758564967835306e-05,
+      "loss": 4.5302,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757726373084254e-05,
+      "loss": 4.5306,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756887778333202e-05,
+      "loss": 4.5269,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756050821462523e-05,
+      "loss": 4.5135,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755212226711471e-05,
+      "loss": 4.5213,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754373631960419e-05,
+      "loss": 4.5086,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75353667508974e-05,
+      "loss": 4.5149,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.752698080338688e-05,
+      "loss": 4.5172,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751859485587636e-05,
+      "loss": 4.5045,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751020890836584e-05,
+      "loss": 4.5124,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7501839339659046e-05,
+      "loss": 4.5099,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.475570201873779,
+      "eval_runtime": 580.8128,
+      "eval_samples_per_second": 656.995,
+      "eval_steps_per_second": 20.532,
+      "step": 152638
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7493453392148526e-05,
+      "loss": 4.5196,
+      "step": 153088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.748506744463801e-05,
+      "loss": 4.5176,
+      "step": 153600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.747669787593122e-05,
+      "loss": 4.5069,
+      "step": 154112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.5055,
+      "step": 154624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.5059,
+      "step": 155136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745154003339966e-05,
+      "loss": 4.4882,
+      "step": 155648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.744315408588914e-05,
+      "loss": 4.5031,
+      "step": 156160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.743476813837862e-05,
+      "loss": 4.5013,
+      "step": 156672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7426382190868095e-05,
+      "loss": 4.4868,
+      "step": 157184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.741801262216131e-05,
+      "loss": 4.5065,
+      "step": 157696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.740962667465079e-05,
+      "loss": 4.5106,
+      "step": 158208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7401240727140264e-05,
+      "loss": 4.495,
+      "step": 158720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.739285477962975e-05,
+      "loss": 4.4883,
+      "step": 159232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.738446883211923e-05,
+      "loss": 4.4931,
+      "step": 159744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.737608288460871e-05,
+      "loss": 4.4957,
+      "step": 160256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.736769693709819e-05,
+      "loss": 4.4908,
+      "step": 160768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735931098958767e-05,
+      "loss": 4.4859,
+      "step": 161280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4854,
+      "step": 161792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4861,
+      "step": 162304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.4893,
+      "step": 162816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.4899,
+      "step": 163328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4921,
+      "step": 163840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4912,
+      "step": 164352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4826,
+      "step": 164864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4868,
+      "step": 165376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4839,
+      "step": 165888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.4861,
+      "step": 166400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.4695,
+      "step": 166912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4822,
+      "step": 167424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4684,
+      "step": 167936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.4734,
+      "step": 168448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4844,
+      "step": 168960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4822,
+      "step": 169472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.721681539712375e-05,
+      "loss": 4.4693,
+      "step": 169984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4644,
+      "step": 170496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4835,
+      "step": 171008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.719165755459219e-05,
+      "loss": 4.4654,
+      "step": 171520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.71832879858854e-05,
+      "loss": 4.4729,
+      "step": 172032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.717490203837488e-05,
+      "loss": 4.4688,
+      "step": 172544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.716651609086437e-05,
+      "loss": 4.4536,
+      "step": 173056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.715813014335385e-05,
+      "loss": 4.4532,
+      "step": 173568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.4529,
+      "step": 174080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714137462713654e-05,
+      "loss": 4.4676,
+      "step": 174592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4585,
+      "step": 175104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.712461911091923e-05,
+      "loss": 4.4585,
+      "step": 175616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.711623316340871e-05,
+      "loss": 4.4521,
+      "step": 176128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.710784721589819e-05,
+      "loss": 4.455,
+      "step": 176640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7099461268387667e-05,
+      "loss": 4.47,
+      "step": 177152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7091075320877147e-05,
+      "loss": 4.451,
+      "step": 177664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7082689373366627e-05,
+      "loss": 4.4511,
+      "step": 178176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7074303425856106e-05,
+      "loss": 4.4458,
+      "step": 178688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7065917478345586e-05,
+      "loss": 4.4685,
+      "step": 179200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.70575479096388e-05,
+      "loss": 4.4375,
+      "step": 179712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704916196212828e-05,
+      "loss": 4.4592,
+      "step": 180224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7040776014617755e-05,
+      "loss": 4.4375,
+      "step": 180736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7032390067107235e-05,
+      "loss": 4.4499,
+      "step": 181248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7024004119596715e-05,
+      "loss": 4.4518,
+      "step": 181760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4399,
+      "step": 182272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4454,
+      "step": 182784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6998862655868884e-05,
+      "loss": 4.4386,
+      "step": 183296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6990476708358364e-05,
+      "loss": 4.4234,
+      "step": 183808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4425,
+      "step": 184320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4416,
+      "step": 184832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6965318865826804e-05,
+      "loss": 4.4332,
+      "step": 185344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.695693291831629e-05,
+      "loss": 4.4333,
+      "step": 185856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694854697080577e-05,
+      "loss": 4.4328,
+      "step": 186368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694016102329525e-05,
+      "loss": 4.416,
+      "step": 186880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.693177507578473e-05,
+      "loss": 4.4439,
+      "step": 187392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.692338912827421e-05,
+      "loss": 4.4246,
+      "step": 187904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.691500318076369e-05,
+      "loss": 4.4283,
+      "step": 188416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.69066336120569e-05,
+      "loss": 4.4389,
+      "step": 188928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.689824766454638e-05,
+      "loss": 4.4266,
+      "step": 189440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.688986171703586e-05,
+      "loss": 4.4369,
+      "step": 189952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.688147576952534e-05,
+      "loss": 4.4458,
+      "step": 190464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.687310620081855e-05,
+      "loss": 4.4315,
+      "step": 190976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4403,
+      "step": 191488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6856350684601245e-05,
+      "loss": 4.421,
+      "step": 192000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6847964737090725e-05,
+      "loss": 4.4132,
+      "step": 192512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6839578789580204e-05,
+      "loss": 4.4257,
+      "step": 193024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6831192842069684e-05,
+      "loss": 4.433,
+      "step": 193536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6822823273362894e-05,
+      "loss": 4.4239,
+      "step": 194048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6814437325852373e-05,
+      "loss": 4.4159,
+      "step": 194560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6806051378341853e-05,
+      "loss": 4.4077,
+      "step": 195072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6797665430831333e-05,
+      "loss": 4.4131,
+      "step": 195584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678927948332081e-05,
+      "loss": 4.4144,
+      "step": 196096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6780893535810287e-05,
+      "loss": 4.4198,
+      "step": 196608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6772507588299766e-05,
+      "loss": 4.4269,
+      "step": 197120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6764121640789246e-05,
+      "loss": 4.4257,
+      "step": 197632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6755735693278726e-05,
+      "loss": 4.4144,
+      "step": 198144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6747349745768206e-05,
+      "loss": 4.4102,
+      "step": 198656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673896379825769e-05,
+      "loss": 4.4057,
+      "step": 199168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673057785074717e-05,
+      "loss": 4.411,
+      "step": 199680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.672222466084412e-05,
+      "loss": 4.4095,
+      "step": 200192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.671383871333359e-05,
+      "loss": 4.4066,
+      "step": 200704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.670545276582307e-05,
+      "loss": 4.4037,
+      "step": 201216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.669706681831255e-05,
+      "loss": 4.4139,
+      "step": 201728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668868087080203e-05,
+      "loss": 4.4117,
+      "step": 202240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668031130209524e-05,
+      "loss": 4.4108,
+      "step": 202752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.667192535458472e-05,
+      "loss": 4.4111,
+      "step": 203264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.66635394070742e-05,
+      "loss": 4.4153,
+      "step": 203776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.665515345956368e-05,
+      "loss": 4.4077,
+      "step": 204288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6646783890856896e-05,
+      "loss": 4.4005,
+      "step": 204800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.4196,
+      "step": 205312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6630011995835856e-05,
+      "loss": 4.3968,
+      "step": 205824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6621626048325336e-05,
+      "loss": 4.4186,
+      "step": 206336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6613240100814816e-05,
+      "loss": 4.3909,
+      "step": 206848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6604854153304296e-05,
+      "loss": 4.3842,
+      "step": 207360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6596468205793776e-05,
+      "loss": 4.3936,
+      "step": 207872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6588082258283256e-05,
+      "loss": 4.4058,
+      "step": 208384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6579712689576465e-05,
+      "loss": 4.3966,
+      "step": 208896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6571326742065945e-05,
+      "loss": 4.3838,
+      "step": 209408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6562940794555425e-05,
+      "loss": 4.3827,
+      "step": 209920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6554554847044905e-05,
+      "loss": 4.3919,
+      "step": 210432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6546185278338114e-05,
+      "loss": 4.406,
+      "step": 210944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.65377993308276e-05,
+      "loss": 4.3915,
+      "step": 211456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652941338331708e-05,
+      "loss": 4.4,
+      "step": 211968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652102743580656e-05,
+      "loss": 4.3953,
+      "step": 212480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.651264148829604e-05,
+      "loss": 4.3924,
+      "step": 212992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.650425554078552e-05,
+      "loss": 4.3948,
+      "step": 213504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.649588597207873e-05,
+      "loss": 4.386,
+      "step": 214016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.648750002456821e-05,
+      "loss": 4.3865,
+      "step": 214528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647911407705769e-05,
+      "loss": 4.3941,
+      "step": 215040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647072812954717e-05,
+      "loss": 4.3791,
+      "step": 215552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.646234218203665e-05,
+      "loss": 4.3858,
+      "step": 216064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.645395623452612e-05,
+      "loss": 4.3886,
+      "step": 216576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.3922,
+      "step": 217088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.3756,
+      "step": 217600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.64288147707983e-05,
+      "loss": 4.377,
+      "step": 218112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.642042882328778e-05,
+      "loss": 4.4018,
+      "step": 218624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6412059254580994e-05,
+      "loss": 4.378,
+      "step": 219136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6403673307070474e-05,
+      "loss": 4.3905,
+      "step": 219648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.639528735955995e-05,
+      "loss": 4.3833,
+      "step": 220160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.638690141204943e-05,
+      "loss": 4.3897,
+      "step": 220672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637851546453891e-05,
+      "loss": 4.3791,
+      "step": 221184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637012951702839e-05,
+      "loss": 4.378,
+      "step": 221696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.3854,
+      "step": 222208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6353374000811076e-05,
+      "loss": 4.3759,
+      "step": 222720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6344988053300556e-05,
+      "loss": 4.3895,
+      "step": 223232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6336602105790036e-05,
+      "loss": 4.3863,
+      "step": 223744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.632823253708325e-05,
+      "loss": 4.3843,
+      "step": 224256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.631984658957273e-05,
+      "loss": 4.3792,
+      "step": 224768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.631146064206221e-05,
+      "loss": 4.3718,
+      "step": 225280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.630307469455169e-05,
+      "loss": 4.3783,
+      "step": 225792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.629468874704117e-05,
+      "loss": 4.3663,
+      "step": 226304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.628630279953065e-05,
+      "loss": 4.3711,
+      "step": 226816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.627791685202013e-05,
+      "loss": 4.3718,
+      "step": 227328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626953090450961e-05,
+      "loss": 4.3645,
+      "step": 227840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626114495699909e-05,
+      "loss": 4.3696,
+      "step": 228352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.625275900948857e-05,
+      "loss": 4.3688,
+      "step": 228864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.342700958251953,
+      "eval_runtime": 554.1608,
+      "eval_samples_per_second": 688.593,
+      "eval_steps_per_second": 21.519,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624437306197805e-05,
+      "loss": 4.3768,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623598711446753e-05,
+      "loss": 4.3819,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622760116695701e-05,
+      "loss": 4.3674,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6219215219446484e-05,
+      "loss": 4.3656,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621082927193597e-05,
+      "loss": 4.3698,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620244332442545e-05,
+      "loss": 4.3501,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619405737691493e-05,
+      "loss": 4.3649,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618567142940441e-05,
+      "loss": 4.3663,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617728548189389e-05,
+      "loss": 4.3541,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616889953438337e-05,
+      "loss": 4.369,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616051358687285e-05,
+      "loss": 4.3763,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615212763936233e-05,
+      "loss": 4.3564,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614374169185181e-05,
+      "loss": 4.3548,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613535574434129e-05,
+      "loss": 4.3592,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612696979683077e-05,
+      "loss": 4.362,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611860022812398e-05,
+      "loss": 4.3564,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611021428061346e-05,
+      "loss": 4.3529,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610182833310294e-05,
+      "loss": 4.3531,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.609344238559242e-05,
+      "loss": 4.3553,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60850564380819e-05,
+      "loss": 4.3616,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607667049057138e-05,
+      "loss": 4.3574,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606828454306086e-05,
+      "loss": 4.362,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6059914974354075e-05,
+      "loss": 4.3602,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6051529026843555e-05,
+      "loss": 4.3544,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604314307933303e-05,
+      "loss": 4.3517,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603475713182251e-05,
+      "loss": 4.3517,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602637118431199e-05,
+      "loss": 4.3578,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601798523680147e-05,
+      "loss": 4.345,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600959928929095e-05,
+      "loss": 4.3536,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600121334178043e-05,
+      "loss": 4.3409,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599284377307364e-05,
+      "loss": 4.345,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5984457825563124e-05,
+      "loss": 4.3589,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597608825685633e-05,
+      "loss": 4.3536,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596770230934581e-05,
+      "loss": 4.3451,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595931636183529e-05,
+      "loss": 4.341,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.59509467931285e-05,
+      "loss": 4.3541,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594256084561798e-05,
+      "loss": 4.3413,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593417489810746e-05,
+      "loss": 4.3528,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.592578895059694e-05,
+      "loss": 4.3436,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591740300308642e-05,
+      "loss": 4.332,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.59090170555759e-05,
+      "loss": 4.3331,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590063110806538e-05,
+      "loss": 4.3344,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589224516055486e-05,
+      "loss": 4.3441,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588385921304434e-05,
+      "loss": 4.3359,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587547326553382e-05,
+      "loss": 4.3373,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58670873180233e-05,
+      "loss": 4.3334,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585870137051279e-05,
+      "loss": 4.3356,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5850331801806e-05,
+      "loss": 4.3503,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584194585429548e-05,
+      "loss": 4.3303,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583355990678496e-05,
+      "loss": 4.3302,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582517395927444e-05,
+      "loss": 4.3331,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581678801176391e-05,
+      "loss": 4.3439,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580840206425339e-05,
+      "loss": 4.3237,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580001611674287e-05,
+      "loss": 4.3381,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579163016923235e-05,
+      "loss": 4.3206,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.578324422172183e-05,
+      "loss": 4.3281,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577487465301504e-05,
+      "loss": 4.3363,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5766488705504526e-05,
+      "loss": 4.3203,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5758102757994006e-05,
+      "loss": 4.3331,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5749716810483486e-05,
+      "loss": 4.328,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741347241776695e-05,
+      "loss": 4.3084,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5732961294266175e-05,
+      "loss": 4.3234,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724575346755655e-05,
+      "loss": 4.3263,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716189399245135e-05,
+      "loss": 4.3244,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5707819830538344e-05,
+      "loss": 4.3199,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699433883027824e-05,
+      "loss": 4.3203,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691047935517304e-05,
+      "loss": 4.3036,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568267836681051e-05,
+      "loss": 4.33,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.567429241929999e-05,
+      "loss": 4.3113,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.566590647178948e-05,
+      "loss": 4.3172,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565752052427896e-05,
+      "loss": 4.3276,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564913457676844e-05,
+      "loss": 4.3155,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564074862925792e-05,
+      "loss": 4.3255,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56323626817474e-05,
+      "loss": 4.3313,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562399311304061e-05,
+      "loss": 4.3251,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561560716553009e-05,
+      "loss": 4.3301,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560722121801957e-05,
+      "loss": 4.3089,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559883527050905e-05,
+      "loss": 4.3067,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559044932299853e-05,
+      "loss": 4.3155,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.558207975429174e-05,
+      "loss": 4.3283,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557369380678122e-05,
+      "loss": 4.3145,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55653078592707e-05,
+      "loss": 4.3079,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555692191176018e-05,
+      "loss": 4.2981,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548552343053394e-05,
+      "loss": 4.3094,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540166395542874e-05,
+      "loss": 4.306,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553179682683608e-05,
+      "loss": 4.3107,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552341087932556e-05,
+      "loss": 4.3178,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551502493181504e-05,
+      "loss": 4.321,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550663898430452e-05,
+      "loss": 4.3087,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549826941559773e-05,
+      "loss": 4.307,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548988346808721e-05,
+      "loss": 4.2969,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548149752057669e-05,
+      "loss": 4.3036,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547311157306617e-05,
+      "loss": 4.308,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546472562555565e-05,
+      "loss": 4.302,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545633967804513e-05,
+      "loss": 4.2973,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544795373053462e-05,
+      "loss": 4.3121,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54395677830241e-05,
+      "loss": 4.3042,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543118183551358e-05,
+      "loss": 4.3154,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542281226680679e-05,
+      "loss": 4.3045,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541442631929627e-05,
+      "loss": 4.3134,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.540604037178575e-05,
+      "loss": 4.306,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539765442427522e-05,
+      "loss": 4.2988,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389284855568436e-05,
+      "loss": 4.317,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5380898908057916e-05,
+      "loss": 4.2952,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372512960547396e-05,
+      "loss": 4.3167,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536412701303687e-05,
+      "loss": 4.2898,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5355741065526356e-05,
+      "loss": 4.2859,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5347355118015836e-05,
+      "loss": 4.2914,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5338985549309045e-05,
+      "loss": 4.3071,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5330599601798525e-05,
+      "loss": 4.2986,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5322213654288005e-05,
+      "loss": 4.2848,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5313827706777485e-05,
+      "loss": 4.2821,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5305458138070694e-05,
+      "loss": 4.2911,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5297072190560174e-05,
+      "loss": 4.309,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5288686243049654e-05,
+      "loss": 4.2924,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5280300295539134e-05,
+      "loss": 4.3041,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5271914348028614e-05,
+      "loss": 4.2967,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5263528400518094e-05,
+      "loss": 4.2971,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255142453007574e-05,
+      "loss": 4.297,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524677288430079e-05,
+      "loss": 4.2884,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523838693679027e-05,
+      "loss": 4.2915,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523000098927975e-05,
+      "loss": 4.2955,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522161504176923e-05,
+      "loss": 4.2852,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521322909425871e-05,
+      "loss": 4.2889,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520485952555192e-05,
+      "loss": 4.2955,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51964735780414e-05,
+      "loss": 4.2928,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518808763053088e-05,
+      "loss": 4.2841,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517970168302036e-05,
+      "loss": 4.2816,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517131573550984e-05,
+      "loss": 4.3032,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516292978799932e-05,
+      "loss": 4.2848,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515456021929253e-05,
+      "loss": 4.2973,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514617427178201e-05,
+      "loss": 4.2899,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513778832427149e-05,
+      "loss": 4.2936,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5129402376760974e-05,
+      "loss": 4.2853,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5121016429250454e-05,
+      "loss": 4.2873,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5112630481739934e-05,
+      "loss": 4.2904,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510424453422941e-05,
+      "loss": 4.2848,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509589134432635e-05,
+      "loss": 4.2984,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508750539681583e-05,
+      "loss": 4.2871,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507911944930531e-05,
+      "loss": 4.2933,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507073350179479e-05,
+      "loss": 4.2916,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506234755428427e-05,
+      "loss": 4.2791,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505396160677375e-05,
+      "loss": 4.2859,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504557565926323e-05,
+      "loss": 4.2771,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503718971175271e-05,
+      "loss": 4.2805,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502882014304593e-05,
+      "loss": 4.2792,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502043419553541e-05,
+      "loss": 4.2783,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501204824802488e-05,
+      "loss": 4.2731,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500366230051436e-05,
+      "loss": 4.2817,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.261253356933594,
+      "eval_runtime": 535.2839,
+      "eval_samples_per_second": 712.876,
+      "eval_steps_per_second": 22.278,
+      "step": 305276
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.499527635300384e-05,
+      "loss": 4.2865,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.498689040549332e-05,
+      "loss": 4.2884,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497852083678653e-05,
+      "loss": 4.2791,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497013488927601e-05,
+      "loss": 4.2755,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.496174894176549e-05,
+      "loss": 4.2828,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.495336299425497e-05,
+      "loss": 4.2597,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494497704674445e-05,
+      "loss": 4.2793,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4936607478037665e-05,
+      "loss": 4.2741,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4928221530527145e-05,
+      "loss": 4.2683,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4919851961820354e-05,
+      "loss": 4.2798,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4911466014309834e-05,
+      "loss": 4.2875,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4903080066799314e-05,
+      "loss": 4.2676,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4894694119288794e-05,
+      "loss": 4.2711,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4886308171778274e-05,
+      "loss": 4.2717,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4877922224267754e-05,
+      "loss": 4.2768,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4869536276757234e-05,
+      "loss": 4.2657,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4861150329246714e-05,
+      "loss": 4.266,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4852764381736194e-05,
+      "loss": 4.2659,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4844378434225674e-05,
+      "loss": 4.267,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4835992486715154e-05,
+      "loss": 4.2779,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4827606539204634e-05,
+      "loss": 4.2693,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481923697049785e-05,
+      "loss": 4.2777,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481085102298733e-05,
+      "loss": 4.2783,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.480246507547681e-05,
+      "loss": 4.2689,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.479407912796629e-05,
+      "loss": 4.2651,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2665,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.477732361174898e-05,
+      "loss": 4.2734,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476893766423846e-05,
+      "loss": 4.2627,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476055171672794e-05,
+      "loss": 4.2666,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475218214802115e-05,
+      "loss": 4.2651,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474379620051063e-05,
+      "loss": 4.2566,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473541025300011e-05,
+      "loss": 4.2719,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472704068429332e-05,
+      "loss": 4.2737,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4718654736782804e-05,
+      "loss": 4.2657,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4710268789272283e-05,
+      "loss": 4.2581,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470189922056549e-05,
+      "loss": 4.2673,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469351327305497e-05,
+      "loss": 4.2668,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468512732554445e-05,
+      "loss": 4.2661,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467674137803393e-05,
+      "loss": 4.2625,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466835543052341e-05,
+      "loss": 4.2522,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.465996948301289e-05,
+      "loss": 4.2503,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4651583535502366e-05,
+      "loss": 4.2516,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4643197587991846e-05,
+      "loss": 4.2631,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4634811640481325e-05,
+      "loss": 4.2571,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4626425692970805e-05,
+      "loss": 4.2543,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618039745460285e-05,
+      "loss": 4.2545,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609653797949765e-05,
+      "loss": 4.2549,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460128422924298e-05,
+      "loss": 4.271,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.459289828173246e-05,
+      "loss": 4.2514,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.458451233422194e-05,
+      "loss": 4.2495,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.457612638671142e-05,
+      "loss": 4.2563,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.456775681800463e-05,
+      "loss": 4.2615,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455937087049411e-05,
+      "loss": 4.2475,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455098492298359e-05,
+      "loss": 4.2546,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.454259897547307e-05,
+      "loss": 4.2458,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.2511,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452584345925576e-05,
+      "loss": 4.256,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451745751174524e-05,
+      "loss": 4.2391,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450907156423472e-05,
+      "loss": 4.2586,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4500685616724206e-05,
+      "loss": 4.2466,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4492316048017415e-05,
+      "loss": 4.236,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4483930100506895e-05,
+      "loss": 4.2431,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4475544152996375e-05,
+      "loss": 4.2492,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4467158205485855e-05,
+      "loss": 4.2527,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4458772257975335e-05,
+      "loss": 4.2419,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4450386310464815e-05,
+      "loss": 4.2471,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4442000362954295e-05,
+      "loss": 4.2235,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4433614415443775e-05,
+      "loss": 4.2536,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4425244846736984e-05,
+      "loss": 4.2372,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.441687527803019e-05,
+      "loss": 4.2404,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440848933051967e-05,
+      "loss": 4.2513,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440010338300916e-05,
+      "loss": 4.242,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.439171743549864e-05,
+      "loss": 4.246,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.438333148798812e-05,
+      "loss": 4.2547,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43749455404776e-05,
+      "loss": 4.2506,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.436655959296708e-05,
+      "loss": 4.2522,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435817364545655e-05,
+      "loss": 4.2374,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434980407674977e-05,
+      "loss": 4.2314,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434141812923925e-05,
+      "loss": 4.2423,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433304856053246e-05,
+      "loss": 4.2516,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432466261302194e-05,
+      "loss": 4.2404,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431627666551142e-05,
+      "loss": 4.2384,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43078907180009e-05,
+      "loss": 4.2226,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429950477049038e-05,
+      "loss": 4.2349,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429111882297986e-05,
+      "loss": 4.231,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428273287546934e-05,
+      "loss": 4.242,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.427434692795882e-05,
+      "loss": 4.2392,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4265977359252026e-05,
+      "loss": 4.2501,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4257591411741506e-05,
+      "loss": 4.233,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4249205464230986e-05,
+      "loss": 4.2383,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4240819516720466e-05,
+      "loss": 4.2233,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4232433569209946e-05,
+      "loss": 4.2317,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4224047621699426e-05,
+      "loss": 4.2364,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4215678052992635e-05,
+      "loss": 4.2269,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4207292105482115e-05,
+      "loss": 4.2275,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4198906157971595e-05,
+      "loss": 4.2411,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419052021046108e-05,
+      "loss": 4.2279,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.418213426295056e-05,
+      "loss": 4.2489,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.417374831544004e-05,
+      "loss": 4.2298,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.416536236792952e-05,
+      "loss": 4.2416,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4156976420419e-05,
+      "loss": 4.2313,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414860685171221e-05,
+      "loss": 4.2293,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414023728300542e-05,
+      "loss": 4.2468,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.41318513354949e-05,
+      "loss": 4.2257,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.412346538798438e-05,
+      "loss": 4.2446,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411507944047386e-05,
+      "loss": 4.2225,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.410669349296334e-05,
+      "loss": 4.2194,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409832392425655e-05,
+      "loss": 4.2202,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408993797674603e-05,
+      "loss": 4.2346,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4081552029235515e-05,
+      "loss": 4.2308,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4073166081724995e-05,
+      "loss": 4.2189,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4064780134214475e-05,
+      "loss": 4.2119,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4056394186703955e-05,
+      "loss": 4.2224,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4048008239193435e-05,
+      "loss": 4.2388,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4039622291682915e-05,
+      "loss": 4.2232,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4031252722976124e-05,
+      "loss": 4.2369,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4022866775465604e-05,
+      "loss": 4.2263,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4014480827955084e-05,
+      "loss": 4.2293,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400609488044456e-05,
+      "loss": 4.2306,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.399772531173777e-05,
+      "loss": 4.22,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398933936422725e-05,
+      "loss": 4.2214,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398095341671673e-05,
+      "loss": 4.2262,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397258384800995e-05,
+      "loss": 4.2188,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396419790049943e-05,
+      "loss": 4.2225,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.395582833179264e-05,
+      "loss": 4.2259,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394744238428212e-05,
+      "loss": 4.2251,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39390564367716e-05,
+      "loss": 4.2162,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393067048926108e-05,
+      "loss": 4.2157,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.392228454175056e-05,
+      "loss": 4.2326,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.391391497304377e-05,
+      "loss": 4.2176,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390552902553325e-05,
+      "loss": 4.2322,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389714307802273e-05,
+      "loss": 4.2272,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3888773509315936e-05,
+      "loss": 4.2242,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388038756180542e-05,
+      "loss": 4.2214,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38720016142949e-05,
+      "loss": 4.22,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.386361566678438e-05,
+      "loss": 4.2235,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.385522971927386e-05,
+      "loss": 4.2185,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3846843771763336e-05,
+      "loss": 4.2333,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3838457824252816e-05,
+      "loss": 4.2247,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3830071876742296e-05,
+      "loss": 4.2277,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3821685929231775e-05,
+      "loss": 4.2254,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3813316360524985e-05,
+      "loss": 4.2174,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3804930413014465e-05,
+      "loss": 4.2174,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3796544465503944e-05,
+      "loss": 4.2132,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3788158517993424e-05,
+      "loss": 4.2139,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377978894928664e-05,
+      "loss": 4.2176,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377140300177612e-05,
+      "loss": 4.2147,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.37630170542656e-05,
+      "loss": 4.2057,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.375463110675508e-05,
+      "loss": 4.2191,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.204667568206787,
+      "eval_runtime": 534.7648,
+      "eval_samples_per_second": 713.568,
+      "eval_steps_per_second": 22.3,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374624515924456e-05,
+      "loss": 4.2284,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373785921173404e-05,
+      "loss": 4.2211,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372947326422352e-05,
+      "loss": 4.2171,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721087316713e-05,
+      "loss": 4.2143,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371270136920248e-05,
+      "loss": 4.2142,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370431542169196e-05,
+      "loss": 4.1992,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369592947418144e-05,
+      "loss": 4.2138,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368754352667092e-05,
+      "loss": 4.2069,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367915757916039e-05,
+      "loss": 4.2114,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367077163164987e-05,
+      "loss": 4.2164,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366238568413936e-05,
+      "loss": 4.2203,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365399973662884e-05,
+      "loss": 4.2051,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364561378911832e-05,
+      "loss": 4.2108,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36372278416078e-05,
+      "loss": 4.2071,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362884189409728e-05,
+      "loss": 4.2143,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362045594658676e-05,
+      "loss": 4.2041,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361206999907624e-05,
+      "loss": 4.203,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360370043036945e-05,
+      "loss": 4.2042,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359531448285893e-05,
+      "loss": 4.2096,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358692853534841e-05,
+      "loss": 4.216,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357854258783789e-05,
+      "loss": 4.2063,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35701730191311e-05,
+      "loss": 4.2191,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356178707162058e-05,
+      "loss": 4.2124,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355340112411006e-05,
+      "loss": 4.2114,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354501517659954e-05,
+      "loss": 4.203,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3536629229089024e-05,
+      "loss": 4.2052,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528243281578504e-05,
+      "loss": 4.2164,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3519857334067984e-05,
+      "loss": 4.2015,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351148776536119e-05,
+      "loss": 4.209,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350310181785067e-05,
+      "loss": 4.201,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349471587034015e-05,
+      "loss": 4.1947,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348632992282963e-05,
+      "loss": 4.2158,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3477943975319106e-05,
+      "loss": 4.2106,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346957440661232e-05,
+      "loss": 4.2095,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34611884591018e-05,
+      "loss": 4.196,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345280251159128e-05,
+      "loss": 4.211,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344443294288449e-05,
+      "loss": 4.2024,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343604699537398e-05,
+      "loss": 4.2113,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342766104786346e-05,
+      "loss": 4.2032,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341927510035294e-05,
+      "loss": 4.1939,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341088915284241e-05,
+      "loss": 4.1916,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340250320533189e-05,
+      "loss": 4.1898,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339411725782137e-05,
+      "loss": 4.208,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338574768911458e-05,
+      "loss": 4.198,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337736174160406e-05,
+      "loss": 4.1923,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336897579409354e-05,
+      "loss": 4.1994,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336058984658302e-05,
+      "loss": 4.1965,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33522038990725e-05,
+      "loss": 4.2103,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334381795156198e-05,
+      "loss": 4.1913,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333543200405146e-05,
+      "loss": 4.1944,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3327046056540946e-05,
+      "loss": 4.1976,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3318676487834155e-05,
+      "loss": 4.2006,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3310306919127365e-05,
+      "loss": 4.1882,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3301920971616844e-05,
+      "loss": 4.1948,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3293535024106324e-05,
+      "loss": 4.1901,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285149076595804e-05,
+      "loss": 4.1966,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3276763129085284e-05,
+      "loss": 4.1932,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268377181574764e-05,
+      "loss": 4.1895,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3259991234064244e-05,
+      "loss": 4.2013,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3251605286553724e-05,
+      "loss": 4.1858,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243219339043204e-05,
+      "loss": 4.1827,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323484977033641e-05,
+      "loss": 4.1846,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32264638228259e-05,
+      "loss": 4.1877,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321807787531538e-05,
+      "loss": 4.203,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320969192780486e-05,
+      "loss": 4.1839,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320130598029434e-05,
+      "loss": 4.1912,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319293641158755e-05,
+      "loss": 4.1707,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318455046407703e-05,
+      "loss": 4.1917,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317618089537024e-05,
+      "loss": 4.1812,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316779494785972e-05,
+      "loss": 4.1833,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31594090003492e-05,
+      "loss": 4.1942,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315102305283868e-05,
+      "loss": 4.187,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314263710532816e-05,
+      "loss": 4.1912,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313426753662137e-05,
+      "loss": 4.1991,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312589796791458e-05,
+      "loss": 4.192,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311751202040406e-05,
+      "loss": 4.1997,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310912607289354e-05,
+      "loss": 4.1841,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310074012538302e-05,
+      "loss": 4.1749,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30923541778725e-05,
+      "loss": 4.1875,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308398460916571e-05,
+      "loss": 4.1967,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307559866165519e-05,
+      "loss": 4.1837,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306721271414467e-05,
+      "loss": 4.1825,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305882676663415e-05,
+      "loss": 4.1698,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305044081912363e-05,
+      "loss": 4.1811,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304205487161311e-05,
+      "loss": 4.1746,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303366892410259e-05,
+      "loss": 4.1865,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302528297659207e-05,
+      "loss": 4.1872,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301691340788529e-05,
+      "loss": 4.1918,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300852746037477e-05,
+      "loss": 4.1822,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300014151286425e-05,
+      "loss": 4.1845,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299175556535372e-05,
+      "loss": 4.17,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29833696178432e-05,
+      "loss": 4.1767,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297498367033268e-05,
+      "loss": 4.1818,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296661410162589e-05,
+      "loss": 4.1752,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295822815411537e-05,
+      "loss": 4.1743,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294984220660485e-05,
+      "loss": 4.1888,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294145625909433e-05,
+      "loss": 4.1739,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293308669038754e-05,
+      "loss": 4.1978,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2924700742877025e-05,
+      "loss": 4.1726,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2916314795366505e-05,
+      "loss": 4.1899,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2907928847855985e-05,
+      "loss": 4.178,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2899542900345465e-05,
+      "loss": 4.1768,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2891156952834945e-05,
+      "loss": 4.196,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2882771005324425e-05,
+      "loss": 4.1706,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2874401436617634e-05,
+      "loss": 4.1915,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2866015489107114e-05,
+      "loss": 4.1688,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2857629541596594e-05,
+      "loss": 4.1677,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849243594086074e-05,
+      "loss": 4.166,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284087402537928e-05,
+      "loss": 4.1802,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283248807786876e-05,
+      "loss": 4.1796,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282410213035824e-05,
+      "loss": 4.1712,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281571618284772e-05,
+      "loss": 4.1582,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280734661414094e-05,
+      "loss": 4.1719,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279896066663042e-05,
+      "loss": 4.1838,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27905747191199e-05,
+      "loss": 4.1784,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278218877160938e-05,
+      "loss": 4.1824,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277380282409886e-05,
+      "loss": 4.1765,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276541687658834e-05,
+      "loss": 4.1772,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275703092907782e-05,
+      "loss": 4.1807,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27486449815673e-05,
+      "loss": 4.1717,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274027541286051e-05,
+      "loss": 4.1702,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273188946534999e-05,
+      "loss": 4.1741,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272350351783947e-05,
+      "loss": 4.1678,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715133949132676e-05,
+      "loss": 4.1717,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270674800162216e-05,
+      "loss": 4.1749,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269836205411164e-05,
+      "loss": 4.173,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268997610660112e-05,
+      "loss": 4.1703,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26815901590906e-05,
+      "loss": 4.1638,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673204211580076e-05,
+      "loss": 4.1805,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2664818264069556e-05,
+      "loss": 4.1661,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656432316559036e-05,
+      "loss": 4.1784,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264806274785225e-05,
+      "loss": 4.1814,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2639676800341725e-05,
+      "loss": 4.1736,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2631290852831205e-05,
+      "loss": 4.1755,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2622904905320685e-05,
+      "loss": 4.1708,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614518957810165e-05,
+      "loss": 4.1721,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606133010299645e-05,
+      "loss": 4.1652,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259776344159286e-05,
+      "loss": 4.183,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258937749408234e-05,
+      "loss": 4.1798,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258099154657182e-05,
+      "loss": 4.1746,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25726055990613e-05,
+      "loss": 4.1739,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256421965155078e-05,
+      "loss": 4.17,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255583370404026e-05,
+      "loss": 4.1712,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254744775652974e-05,
+      "loss": 4.1641,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253906180901922e-05,
+      "loss": 4.1593,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253069224031243e-05,
+      "loss": 4.1722,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252230629280191e-05,
+      "loss": 4.1643,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251392034529139e-05,
+      "loss": 4.155,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250553439778087e-05,
+      "loss": 4.1676,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.164071083068848,
+      "eval_runtime": 556.8816,
+      "eval_samples_per_second": 685.228,
+      "eval_steps_per_second": 21.414,
+      "step": 457914
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.854320272228581e+17,
+  "trial_name": null,
+  "trial_params": null
+}