diff --git "a/checkpoint-2671184/trainer_state.json" "b/checkpoint-2671184/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2671184/trainer_state.json"
@@ -0,0 +1,31607 @@
+{
+  "best_metric": 3.981346607208252,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl/lstm/4/checkpoints/checkpoint-2671184",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 2671184,
+  "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
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497164829074085e-05,
+      "loss": 4.1795,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488778881563565e-05,
+      "loss": 4.1726,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480392934053045e-05,
+      "loss": 4.1683,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472006986542525e-05,
+      "loss": 4.166,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2463621039032005e-05,
+      "loss": 4.1672,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455251470325214e-05,
+      "loss": 4.1548,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2446865522814694e-05,
+      "loss": 4.1647,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2438479575304174e-05,
+      "loss": 4.1585,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430093627793654e-05,
+      "loss": 4.1624,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2421707680283134e-05,
+      "loss": 4.1668,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413321732772614e-05,
+      "loss": 4.1758,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240493578526209e-05,
+      "loss": 4.1526,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239654983775157e-05,
+      "loss": 4.168,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238816389024105e-05,
+      "loss": 4.1594,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379777942730534e-05,
+      "loss": 4.1705,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371391995220014e-05,
+      "loss": 4.153,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363006047709494e-05,
+      "loss": 4.1546,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354620100198974e-05,
+      "loss": 4.1608,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346234152688454e-05,
+      "loss": 4.158,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337848205177934e-05,
+      "loss": 4.1671,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329462257667414e-05,
+      "loss": 4.1604,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232109268896062e-05,
+      "loss": 4.17,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23127067414501e-05,
+      "loss": 4.1673,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230432079393958e-05,
+      "loss": 4.1609,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229593484642906e-05,
+      "loss": 4.1549,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228754889891854e-05,
+      "loss": 4.1578,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227916295140802e-05,
+      "loss": 4.1693,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22707770038975e-05,
+      "loss": 4.157,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226239105638698e-05,
+      "loss": 4.16,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22540214876802e-05,
+      "loss": 4.1582,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224563554016968e-05,
+      "loss": 4.1493,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223724959265915e-05,
+      "loss": 4.1659,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222886364514863e-05,
+      "loss": 4.1654,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222047769763811e-05,
+      "loss": 4.1647,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221210812893133e-05,
+      "loss": 4.1479,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22037221814208e-05,
+      "loss": 4.1636,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219533623391028e-05,
+      "loss": 4.1563,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218695028639976e-05,
+      "loss": 4.1634,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217856433888924e-05,
+      "loss": 4.1579,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217017839137872e-05,
+      "loss": 4.1493,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21617924438682e-05,
+      "loss": 4.1411,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215340649635769e-05,
+      "loss": 4.1419,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214502054884717e-05,
+      "loss": 4.1662,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2136650980140376e-05,
+      "loss": 4.1512,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128265032629856e-05,
+      "loss": 4.1462,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2119879085119336e-05,
+      "loss": 4.1583,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111493137608816e-05,
+      "loss": 4.1503,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103123568902025e-05,
+      "loss": 4.1624,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094737621391505e-05,
+      "loss": 4.1492,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2086351673880985e-05,
+      "loss": 4.1524,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2077965726370465e-05,
+      "loss": 4.1501,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2069596157663674e-05,
+      "loss": 4.155,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061210210153154e-05,
+      "loss": 4.1483,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052824262642634e-05,
+      "loss": 4.1462,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204443831513212e-05,
+      "loss": 4.1475,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20360523676216e-05,
+      "loss": 4.1539,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202766642011108e-05,
+      "loss": 4.1452,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201928047260056e-05,
+      "loss": 4.1438,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201089452509004e-05,
+      "loss": 4.157,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200252495638325e-05,
+      "loss": 4.1438,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199413900887273e-05,
+      "loss": 4.1383,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198575306136221e-05,
+      "loss": 4.1396,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197736711385169e-05,
+      "loss": 4.1433,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19689975451449e-05,
+      "loss": 4.1615,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196061159763438e-05,
+      "loss": 4.1395,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195222565012386e-05,
+      "loss": 4.1466,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194383970261334e-05,
+      "loss": 4.1284,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1935470133906554e-05,
+      "loss": 4.1468,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927084186396034e-05,
+      "loss": 4.1369,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918698238885514e-05,
+      "loss": 4.1417,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191031229137499e-05,
+      "loss": 4.1482,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19019427226682e-05,
+      "loss": 4.1426,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189355677515768e-05,
+      "loss": 4.1491,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188517082764716e-05,
+      "loss": 4.1538,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876784880136636e-05,
+      "loss": 4.1502,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186841531142985e-05,
+      "loss": 4.1541,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186002936391933e-05,
+      "loss": 4.147,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185165979521254e-05,
+      "loss": 4.1302,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184327384770203e-05,
+      "loss": 4.1426,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183488790019151e-05,
+      "loss": 4.1487,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182650195268099e-05,
+      "loss": 4.1454,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181811600517046e-05,
+      "loss": 4.1424,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180974643646368e-05,
+      "loss": 4.1233,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180136048895316e-05,
+      "loss": 4.1409,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179297454144264e-05,
+      "loss": 4.131,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178458859393211e-05,
+      "loss": 4.1467,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177620264642159e-05,
+      "loss": 4.1427,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176781669891107e-05,
+      "loss": 4.1505,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175943075140055e-05,
+      "loss": 4.1371,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751061182693766e-05,
+      "loss": 4.1414,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1742675235183246e-05,
+      "loss": 4.1297,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734289287672726e-05,
+      "loss": 4.1405,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1725903340162206e-05,
+      "loss": 4.1347,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1717517392651685e-05,
+      "loss": 4.1353,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709131445141165e-05,
+      "loss": 4.1292,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1700745497630645e-05,
+      "loss": 4.1426,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692375928923855e-05,
+      "loss": 4.1338,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1683989981413334e-05,
+      "loss": 4.1579,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1675604033902814e-05,
+      "loss": 4.1319,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667218086392294e-05,
+      "loss": 4.1482,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1658832138881774e-05,
+      "loss": 4.1373,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1650446191371254e-05,
+      "loss": 4.1361,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642060243860734e-05,
+      "loss": 4.1511,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163369067515395e-05,
+      "loss": 4.1359,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162530472764343e-05,
+      "loss": 4.1468,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161691878013291e-05,
+      "loss": 4.1339,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160853283262239e-05,
+      "loss": 4.1226,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160014688511187e-05,
+      "loss": 4.1253,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159176093760135e-05,
+      "loss": 4.14,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158337499009082e-05,
+      "loss": 4.138,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15749890425803e-05,
+      "loss": 4.1339,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156663585267725e-05,
+      "loss": 4.1123,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155824990516673e-05,
+      "loss": 4.13,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154988033645994e-05,
+      "loss": 4.1438,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154149438894942e-05,
+      "loss": 4.1374,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533108441438904e-05,
+      "loss": 4.1424,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524722493928384e-05,
+      "loss": 4.1376,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516336546417864e-05,
+      "loss": 4.1371,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1507950598907344e-05,
+      "loss": 4.1392,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149958103020055e-05,
+      "loss": 4.1334,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149119508269003e-05,
+      "loss": 4.1273,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148280913517951e-05,
+      "loss": 4.1333,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147442318766899e-05,
+      "loss": 4.1296,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146603724015847e-05,
+      "loss": 4.1325,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457651292647946e-05,
+      "loss": 4.1366,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449265345137426e-05,
+      "loss": 4.1324,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1440879397626906e-05,
+      "loss": 4.128,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432493450116386e-05,
+      "loss": 4.1261,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424107502605866e-05,
+      "loss": 4.1396,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141572155509535e-05,
+      "loss": 4.1261,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140733560758483e-05,
+      "loss": 4.1408,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139896603887804e-05,
+      "loss": 4.1443,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139058009136752e-05,
+      "loss": 4.1328,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138221052266073e-05,
+      "loss": 4.136,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137382457515021e-05,
+      "loss": 4.1329,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136543862763969e-05,
+      "loss": 4.1302,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135705268012917e-05,
+      "loss": 4.1267,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134866673261865e-05,
+      "loss": 4.144,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134028078510813e-05,
+      "loss": 4.1404,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133189483759761e-05,
+      "loss": 4.1378,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132350889008709e-05,
+      "loss": 4.136,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315139321380306e-05,
+      "loss": 4.1283,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306753373869786e-05,
+      "loss": 4.1301,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298367426359266e-05,
+      "loss": 4.1274,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1289997857652475e-05,
+      "loss": 4.1204,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281611910141955e-05,
+      "loss": 4.1365,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273225962631435e-05,
+      "loss": 4.1263,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1264840015120915e-05,
+      "loss": 4.1168,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1256454067610395e-05,
+      "loss": 4.1315,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.133627414703369,
+      "eval_runtime": 530.381,
+      "eval_samples_per_second": 719.466,
+      "eval_steps_per_second": 22.484,
+      "step": 534233
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1248084498903604e-05,
+      "loss": 4.1414,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1239698551393084e-05,
+      "loss": 4.1321,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1231312603882564e-05,
+      "loss": 4.1368,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1222926656372044e-05,
+      "loss": 4.127,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1214540708861524e-05,
+      "loss": 4.1244,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1206154761351004e-05,
+      "loss": 4.1207,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1197768813840484e-05,
+      "loss": 4.1247,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1189382866329964e-05,
+      "loss": 4.122,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1180996918819444e-05,
+      "loss": 4.1231,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1172610971308923e-05,
+      "loss": 4.1242,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1164225023798403e-05,
+      "loss": 4.139,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1155839076287883e-05,
+      "loss": 4.1183,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114745312877736e-05,
+      "loss": 4.1303,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113906718126684e-05,
+      "loss": 4.1162,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113068123375632e-05,
+      "loss": 4.133,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11222952862458e-05,
+      "loss": 4.116,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111390933873528e-05,
+      "loss": 4.1184,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110552339122476e-05,
+      "loss": 4.1206,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109717020132171e-05,
+      "loss": 4.1204,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108878425381119e-05,
+      "loss": 4.1288,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108039830630067e-05,
+      "loss": 4.1241,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107201235879015e-05,
+      "loss": 4.1362,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106362641127963e-05,
+      "loss": 4.1254,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105524046376911e-05,
+      "loss": 4.1264,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104685451625859e-05,
+      "loss": 4.1208,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103846856874807e-05,
+      "loss": 4.1218,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103008262123755e-05,
+      "loss": 4.1272,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102169667372702e-05,
+      "loss": 4.1196,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10133107262165e-05,
+      "loss": 4.1257,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100494115750972e-05,
+      "loss": 4.1202,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09965552099992e-05,
+      "loss": 4.1142,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098816926248868e-05,
+      "loss": 4.1261,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097978331497816e-05,
+      "loss": 4.1309,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097141374627137e-05,
+      "loss": 4.125,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963027798760846e-05,
+      "loss": 4.1142,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0954641851250326e-05,
+      "loss": 4.1273,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946255903739806e-05,
+      "loss": 4.1206,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0937869956229286e-05,
+      "loss": 4.1269,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0929484008718766e-05,
+      "loss": 4.1211,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921098061208245e-05,
+      "loss": 4.1156,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0912712113697725e-05,
+      "loss": 4.1069,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0904326166187205e-05,
+      "loss": 4.108,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0895956597480414e-05,
+      "loss": 4.1261,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0887570649969894e-05,
+      "loss": 4.1168,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087920108126311e-05,
+      "loss": 4.1115,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087081513375259e-05,
+      "loss": 4.1215,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086242918624207e-05,
+      "loss": 4.1138,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085404323873155e-05,
+      "loss": 4.1268,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084565729122103e-05,
+      "loss": 4.1122,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083727134371051e-05,
+      "loss": 4.1195,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082888539619999e-05,
+      "loss": 4.1115,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08205158274932e-05,
+      "loss": 4.1199,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081212987998268e-05,
+      "loss": 4.1133,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080374393247216e-05,
+      "loss": 4.1078,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079535798496164e-05,
+      "loss": 4.1158,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078697203745112e-05,
+      "loss": 4.1149,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07785860899406e-05,
+      "loss": 4.1107,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0770216521233815e-05,
+      "loss": 4.1109,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0761830573723295e-05,
+      "loss": 4.1234,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0753444626212775e-05,
+      "loss": 4.102,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745058678702255e-05,
+      "loss": 4.1062,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0736672731191735e-05,
+      "loss": 4.1045,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072828678368121e-05,
+      "loss": 4.1094,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071990083617069e-05,
+      "loss": 4.1234,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071151488866017e-05,
+      "loss": 4.1058,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070312894114965e-05,
+      "loss": 4.1117,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069475937244286e-05,
+      "loss": 4.096,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068638980373607e-05,
+      "loss": 4.1105,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067800385622555e-05,
+      "loss": 4.1026,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066961790871503e-05,
+      "loss": 4.1097,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066123196120451e-05,
+      "loss": 4.1127,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065284601369399e-05,
+      "loss": 4.1084,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064447644498721e-05,
+      "loss": 4.1122,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063609049747668e-05,
+      "loss": 4.1213,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062770454996616e-05,
+      "loss": 4.1122,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061931860245564e-05,
+      "loss": 4.1195,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061093265494512e-05,
+      "loss": 4.1101,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06025467074346e-05,
+      "loss": 4.0985,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059417713872781e-05,
+      "loss": 4.108,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0585807570021026e-05,
+      "loss": 4.1142,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0577421622510506e-05,
+      "loss": 4.1132,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0569035674999986e-05,
+      "loss": 4.107,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0560649727489466e-05,
+      "loss": 4.0924,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0552263779978946e-05,
+      "loss": 4.1027,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0543877832468426e-05,
+      "loss": 4.0989,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0535491884957906e-05,
+      "loss": 4.1117,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0527105937447386e-05,
+      "loss": 4.1067,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0518719989936866e-05,
+      "loss": 4.1144,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0510334042426346e-05,
+      "loss": 4.1075,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0501964473719555e-05,
+      "loss": 4.1035,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0493578526209035e-05,
+      "loss": 4.0995,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0485192578698515e-05,
+      "loss": 4.1046,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0476806631187995e-05,
+      "loss": 4.0985,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0468420683677475e-05,
+      "loss": 4.1012,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460034736166955e-05,
+      "loss": 4.0981,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0451648788656435e-05,
+      "loss": 4.1091,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443262841145915e-05,
+      "loss": 4.1017,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0434876893635395e-05,
+      "loss": 4.1251,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042650732492861e-05,
+      "loss": 4.0949,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041812137741809e-05,
+      "loss": 4.115,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040973542990757e-05,
+      "loss": 4.1075,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0401349482397044e-05,
+      "loss": 4.1007,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039297991369026e-05,
+      "loss": 4.1114,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038461034498347e-05,
+      "loss": 4.1052,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037622439747295e-05,
+      "loss": 4.1095,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036783844996243e-05,
+      "loss": 4.1011,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035945250245191e-05,
+      "loss": 4.0909,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035106655494139e-05,
+      "loss": 4.0941,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034268060743087e-05,
+      "loss": 4.1046,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0334311038724084e-05,
+      "loss": 4.1063,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0325925091213564e-05,
+      "loss": 4.0988,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0317539143703044e-05,
+      "loss": 4.0794,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030915319619252e-05,
+      "loss": 4.0974,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0300767248682e-05,
+      "loss": 4.1102,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029238130117148e-05,
+      "loss": 4.1022,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028399535366096e-05,
+      "loss": 4.1129,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027560940615044e-05,
+      "loss": 4.1044,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026722345863992e-05,
+      "loss": 4.0989,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0258853889933126e-05,
+      "loss": 4.1077,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025050070003008e-05,
+      "loss": 4.1017,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024211475251956e-05,
+      "loss": 4.0945,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.023372880500904e-05,
+      "loss": 4.1012,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022534285749852e-05,
+      "loss": 4.0969,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021695690998799e-05,
+      "loss": 4.0971,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020857096247747e-05,
+      "loss": 4.0983,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020018501496695e-05,
+      "loss": 4.1053,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019179906745643e-05,
+      "loss": 4.0962,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018341311994591e-05,
+      "loss": 4.0904,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017502717243539e-05,
+      "loss": 4.1099,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016664122492487e-05,
+      "loss": 4.0902,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015825527741435e-05,
+      "loss": 4.1067,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014988570870756e-05,
+      "loss": 4.1129,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0141499761197047e-05,
+      "loss": 4.103,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0133113813686527e-05,
+      "loss": 4.1006,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0124727866176006e-05,
+      "loss": 4.0989,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0116358297469216e-05,
+      "loss": 4.0993,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0107972349958696e-05,
+      "loss": 4.095,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0099586402448175e-05,
+      "loss": 4.1138,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0091200454937655e-05,
+      "loss": 4.1066,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0082830886230865e-05,
+      "loss": 4.1079,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0074444938720344e-05,
+      "loss": 4.101,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0066058991209824e-05,
+      "loss": 4.097,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0057673043699304e-05,
+      "loss": 4.1012,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0049287096188784e-05,
+      "loss": 4.0914,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004093390628573e-05,
+      "loss": 4.0913,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003254795877521e-05,
+      "loss": 4.1022,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002416201126469e-05,
+      "loss": 4.0949,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001577606375417e-05,
+      "loss": 4.0826,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000739011624365e-05,
+      "loss": 4.1017,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.110222339630127,
+      "eval_runtime": 528.7786,
+      "eval_samples_per_second": 721.646,
+      "eval_steps_per_second": 22.552,
+      "step": 610552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999902054753686e-05,
+      "loss": 4.0969,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999063460002634e-05,
+      "loss": 4.1058,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998224865251582e-05,
+      "loss": 4.1019,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99738627050053e-05,
+      "loss": 4.0946,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996547675749478e-05,
+      "loss": 4.0933,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995710718878799e-05,
+      "loss": 4.0908,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994872124127747e-05,
+      "loss": 4.0915,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9940335293766954e-05,
+      "loss": 4.0893,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9931949346256434e-05,
+      "loss": 4.09,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9923563398745914e-05,
+      "loss": 4.0954,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991519383003912e-05,
+      "loss": 4.1054,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99068078825286e-05,
+      "loss": 4.0879,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989842193501808e-05,
+      "loss": 4.0988,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989003598750756e-05,
+      "loss": 4.0835,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988165003999704e-05,
+      "loss": 4.1057,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987326409248652e-05,
+      "loss": 4.0811,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9864878144975996e-05,
+      "loss": 4.0881,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9856492197465476e-05,
+      "loss": 4.0904,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984813900756242e-05,
+      "loss": 4.0905,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983975306005191e-05,
+      "loss": 4.0971,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983138349134512e-05,
+      "loss": 4.0925,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9822997543834597e-05,
+      "loss": 4.1076,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9814611596324077e-05,
+      "loss": 4.0938,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9806225648813557e-05,
+      "loss": 4.0957,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9797839701303036e-05,
+      "loss": 4.0866,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9789453753792516e-05,
+      "loss": 4.0952,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781067806281996e-05,
+      "loss": 4.0953,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.977268185877147e-05,
+      "loss": 4.089,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976429591126095e-05,
+      "loss": 4.0945,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.975590996375043e-05,
+      "loss": 4.0905,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974752401623991e-05,
+      "loss": 4.0805,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973913806872939e-05,
+      "loss": 4.0967,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9730768500022605e-05,
+      "loss": 4.0999,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9722382552512085e-05,
+      "loss": 4.093,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9713996605001565e-05,
+      "loss": 4.0822,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9705610657491045e-05,
+      "loss": 4.0973,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9697241088784254e-05,
+      "loss": 4.0893,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9688855141273734e-05,
+      "loss": 4.0926,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9680469193763214e-05,
+      "loss": 4.0931,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9672083246252694e-05,
+      "loss": 4.0834,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9663697298742174e-05,
+      "loss": 4.0781,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9655311351231654e-05,
+      "loss": 4.0748,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9646925403721134e-05,
+      "loss": 4.0956,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9638539456210614e-05,
+      "loss": 4.0861,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963016988750382e-05,
+      "loss": 4.081,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962180031879704e-05,
+      "loss": 4.0943,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961341437128652e-05,
+      "loss": 4.0767,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9605028423776e-05,
+      "loss": 4.0982,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959664247626548e-05,
+      "loss": 4.0835,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958825652875496e-05,
+      "loss": 4.0893,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957987058124444e-05,
+      "loss": 4.0832,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957148463373392e-05,
+      "loss": 4.0838,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95630986862234e-05,
+      "loss": 4.0898,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955472911751661e-05,
+      "loss": 4.0737,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954635954880982e-05,
+      "loss": 4.0889,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95379736012993e-05,
+      "loss": 4.0834,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952958765378878e-05,
+      "loss": 4.0829,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9521201706278263e-05,
+      "loss": 4.0827,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.951281575876774e-05,
+      "loss": 4.093,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950442981125722e-05,
+      "loss": 4.0688,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94960438637467e-05,
+      "loss": 4.0794,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948765791623618e-05,
+      "loss": 4.0752,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947928834752939e-05,
+      "loss": 4.0805,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947090240001887e-05,
+      "loss": 4.0895,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946251645250835e-05,
+      "loss": 4.0782,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945413050499783e-05,
+      "loss": 4.0868,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9445744557487305e-05,
+      "loss": 4.06,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943737498878052e-05,
+      "loss": 4.0859,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942898904127e-05,
+      "loss": 4.0734,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942060309375948e-05,
+      "loss": 4.0781,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941221714624896e-05,
+      "loss": 4.0796,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940384757754218e-05,
+      "loss": 4.0816,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939546163003166e-05,
+      "loss": 4.0809,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9387092061324866e-05,
+      "loss": 4.0954,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9378706113814346e-05,
+      "loss": 4.0814,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9370320166303826e-05,
+      "loss": 4.0898,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9361934218793306e-05,
+      "loss": 4.0847,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.935354827128278e-05,
+      "loss": 4.0711,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9345178702575995e-05,
+      "loss": 4.0727,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9336792755065475e-05,
+      "loss": 4.0844,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9328406807554955e-05,
+      "loss": 4.089,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9320020860044435e-05,
+      "loss": 4.0829,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9311634912533915e-05,
+      "loss": 4.0619,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9303248965023395e-05,
+      "loss": 4.0702,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9294863017512875e-05,
+      "loss": 4.0727,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9286477070002355e-05,
+      "loss": 4.08,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9278107501295564e-05,
+      "loss": 4.0821,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926973793258878e-05,
+      "loss": 4.0838,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926135198507825e-05,
+      "loss": 4.0771,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925296603756773e-05,
+      "loss": 4.0761,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924458009005721e-05,
+      "loss": 4.0727,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923619414254669e-05,
+      "loss": 4.0738,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922784095264364e-05,
+      "loss": 4.0645,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9219455005133124e-05,
+      "loss": 4.0804,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9211069057622604e-05,
+      "loss": 4.0714,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920268311011208e-05,
+      "loss": 4.0787,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919429716260156e-05,
+      "loss": 4.0732,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918591121509104e-05,
+      "loss": 4.0928,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917752526758052e-05,
+      "loss": 4.0693,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916913932007e-05,
+      "loss": 4.0851,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9160769751363206e-05,
+      "loss": 4.083,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9152383803852686e-05,
+      "loss": 4.0737,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9143997856342166e-05,
+      "loss": 4.0803,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9135611908831646e-05,
+      "loss": 4.0816,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912724234012486e-05,
+      "loss": 4.0823,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911885639261434e-05,
+      "loss": 4.0725,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911047044510382e-05,
+      "loss": 4.0643,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91020844975933e-05,
+      "loss": 4.0684,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909369855008278e-05,
+      "loss": 4.0765,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908531260257226e-05,
+      "loss": 4.0758,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.907692665506174e-05,
+      "loss": 4.0754,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906854070755122e-05,
+      "loss": 4.0503,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90601547600407e-05,
+      "loss": 4.0726,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905176881253018e-05,
+      "loss": 4.0792,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904338286501966e-05,
+      "loss": 4.0762,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903499691750914e-05,
+      "loss": 4.0836,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902662734880235e-05,
+      "loss": 4.0784,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901824140129183e-05,
+      "loss": 4.0685,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900985545378131e-05,
+      "loss": 4.0812,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900146950627079e-05,
+      "loss": 4.0773,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8993099937564007e-05,
+      "loss": 4.0652,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8984713990053487e-05,
+      "loss": 4.0736,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8976328042542966e-05,
+      "loss": 4.0719,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.896794209503244e-05,
+      "loss": 4.0705,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8959572526325656e-05,
+      "loss": 4.0714,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8951186578815135e-05,
+      "loss": 4.0745,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8942800631304615e-05,
+      "loss": 4.0735,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.893441468379409e-05,
+      "loss": 4.0626,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.892602873628357e-05,
+      "loss": 4.0788,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891764278877305e-05,
+      "loss": 4.0647,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890925684126253e-05,
+      "loss": 4.0798,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890087089375201e-05,
+      "loss": 4.0889,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889251770384896e-05,
+      "loss": 4.0731,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.888413175633844e-05,
+      "loss": 4.075,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887574580882791e-05,
+      "loss": 4.0692,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886735986131739e-05,
+      "loss": 4.0734,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885897391380687e-05,
+      "loss": 4.0676,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885058796629635e-05,
+      "loss": 4.0836,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884221839758956e-05,
+      "loss": 4.0842,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883383245007904e-05,
+      "loss": 4.0757,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882544650256852e-05,
+      "loss": 4.0738,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8817060555058e-05,
+      "loss": 4.0758,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880867460754748e-05,
+      "loss": 4.0706,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880028866003696e-05,
+      "loss": 4.0686,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879191909133018e-05,
+      "loss": 4.0647,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878354952262339e-05,
+      "loss": 4.0733,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877516357511287e-05,
+      "loss": 4.0687,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876677762760235e-05,
+      "loss": 4.058,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.875839168009183e-05,
+      "loss": 4.0701,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.090645790100098,
+      "eval_runtime": 524.3925,
+      "eval_samples_per_second": 727.682,
+      "eval_steps_per_second": 22.741,
+      "step": 686871
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.875000573258131e-05,
+      "loss": 4.0753,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874161978507079e-05,
+      "loss": 4.0798,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873323383756027e-05,
+      "loss": 4.0775,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872484789004975e-05,
+      "loss": 4.0634,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871646194253923e-05,
+      "loss": 4.0718,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870807599502871e-05,
+      "loss": 4.065,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869969004751819e-05,
+      "loss": 4.0632,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8691304100007667e-05,
+      "loss": 4.0661,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8682918152497147e-05,
+      "loss": 4.0613,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8674532204986626e-05,
+      "loss": 4.0712,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8666146257476106e-05,
+      "loss": 4.0744,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8657760309965586e-05,
+      "loss": 4.0611,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8649374362455066e-05,
+      "loss": 4.0731,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8640988414944546e-05,
+      "loss": 4.0569,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8632602467434026e-05,
+      "loss": 4.0745,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8624216519923506e-05,
+      "loss": 4.0541,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8615830572412986e-05,
+      "loss": 4.0622,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8607444624902466e-05,
+      "loss": 4.066,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599075056195675e-05,
+      "loss": 4.0637,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8590689108685155e-05,
+      "loss": 4.0741,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8582303161174635e-05,
+      "loss": 4.0596,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8573917213664115e-05,
+      "loss": 4.0854,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8565531266153595e-05,
+      "loss": 4.0647,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855714531864308e-05,
+      "loss": 4.0681,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854875937113256e-05,
+      "loss": 4.0629,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854038980242577e-05,
+      "loss": 4.0718,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853200385491525e-05,
+      "loss": 4.0682,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852361790740473e-05,
+      "loss": 4.0612,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851523195989421e-05,
+      "loss": 4.0691,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850687876999115e-05,
+      "loss": 4.0681,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849849282248063e-05,
+      "loss": 4.0497,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849010687497011e-05,
+      "loss": 4.0718,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.848172092745959e-05,
+      "loss": 4.0726,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.847333497994907e-05,
+      "loss": 4.0679,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846494903243855e-05,
+      "loss": 4.0569,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8456563084928035e-05,
+      "loss": 4.0719,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8448193516221245e-05,
+      "loss": 4.0644,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8439807568710724e-05,
+      "loss": 4.067,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8431421621200204e-05,
+      "loss": 4.0649,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8423035673689684e-05,
+      "loss": 4.056,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8414649726179164e-05,
+      "loss": 4.0544,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840626377866864e-05,
+      "loss": 4.0484,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839787783115812e-05,
+      "loss": 4.0706,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83894918836476e-05,
+      "loss": 4.0604,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838110593613708e-05,
+      "loss": 4.0537,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8372736367430287e-05,
+      "loss": 4.0664,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836435041991977e-05,
+      "loss": 4.0527,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835596447240925e-05,
+      "loss": 4.0701,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834757852489873e-05,
+      "loss": 4.0635,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833919257738821e-05,
+      "loss": 4.0635,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833080662987769e-05,
+      "loss": 4.059,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832242068236717e-05,
+      "loss": 4.0552,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831405111366038e-05,
+      "loss": 4.0652,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830566516614986e-05,
+      "loss": 4.0461,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829727921863934e-05,
+      "loss": 4.064,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828889327112882e-05,
+      "loss": 4.0533,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82805073236183e-05,
+      "loss": 4.0571,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827213775491151e-05,
+      "loss": 4.054,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826375180740099e-05,
+      "loss": 4.0669,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825536585989047e-05,
+      "loss": 4.0506,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824697991237996e-05,
+      "loss": 4.0495,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823859396486944e-05,
+      "loss": 4.0563,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823020801735892e-05,
+      "loss": 4.0487,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82218220698484e-05,
+      "loss": 4.0603,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821345250114161e-05,
+      "loss": 4.0586,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8205066553631087e-05,
+      "loss": 4.0603,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8196680606120567e-05,
+      "loss": 4.0364,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8188294658610047e-05,
+      "loss": 4.0575,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817990871109952e-05,
+      "loss": 4.0492,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171522763589e-05,
+      "loss": 4.0506,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816313681607848e-05,
+      "loss": 4.0549,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154767247371695e-05,
+      "loss": 4.0554,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8146381299861175e-05,
+      "loss": 4.0567,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8137995352350655e-05,
+      "loss": 4.0706,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8129609404840135e-05,
+      "loss": 4.0557,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8121223457329615e-05,
+      "loss": 4.0624,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8112837509819095e-05,
+      "loss": 4.0569,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8104467941112304e-05,
+      "loss": 4.0497,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8096081993601784e-05,
+      "loss": 4.0488,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8087712424894993e-05,
+      "loss": 4.0562,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807932647738447e-05,
+      "loss": 4.06,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807094052987395e-05,
+      "loss": 4.0627,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806255458236343e-05,
+      "loss": 4.0354,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805416863485291e-05,
+      "loss": 4.0427,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804578268734239e-05,
+      "loss": 4.0479,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803739673983187e-05,
+      "loss": 4.055,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802901079232136e-05,
+      "loss": 4.0522,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802064122361457e-05,
+      "loss": 4.0637,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801225527610405e-05,
+      "loss": 4.0514,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800386932859353e-05,
+      "loss": 4.0518,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799548338108301e-05,
+      "loss": 4.0464,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798711381237622e-05,
+      "loss": 4.0484,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79787278648657e-05,
+      "loss": 4.0402,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797034191735518e-05,
+      "loss": 4.0519,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796195596984466e-05,
+      "loss": 4.0502,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795357002233414e-05,
+      "loss": 4.0505,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794518407482362e-05,
+      "loss": 4.0525,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79367981273131e-05,
+      "loss": 4.0644,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792841217980258e-05,
+      "loss": 4.0459,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920042611095793e-05,
+      "loss": 4.0598,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911673042389e-05,
+      "loss": 4.0568,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.790328709487848e-05,
+      "loss": 4.0522,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789490114736796e-05,
+      "loss": 4.049,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788651519985744e-05,
+      "loss": 4.0574,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787812925234692e-05,
+      "loss": 4.0573,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78697433048364e-05,
+      "loss": 4.0441,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786135735732588e-05,
+      "loss": 4.0433,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7852971409815355e-05,
+      "loss": 4.0416,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784460184110857e-05,
+      "loss": 4.0467,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783621589359805e-05,
+      "loss": 4.0547,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782782994608753e-05,
+      "loss": 4.0481,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781944399857701e-05,
+      "loss": 4.0323,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781107442987023e-05,
+      "loss": 4.0437,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780268848235971e-05,
+      "loss": 4.0526,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779430253484919e-05,
+      "loss": 4.054,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778591658733866e-05,
+      "loss": 4.0622,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7777547018631876e-05,
+      "loss": 4.0518,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7769161071121356e-05,
+      "loss": 4.046,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776077512361083e-05,
+      "loss": 4.0526,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775238917610031e-05,
+      "loss": 4.054,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774400322858979e-05,
+      "loss": 4.046,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773561728107927e-05,
+      "loss": 4.0435,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772723133356875e-05,
+      "loss": 4.0517,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7718845386058236e-05,
+      "loss": 4.0398,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7710475817351445e-05,
+      "loss": 4.0509,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770210624864466e-05,
+      "loss": 4.047,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7693720301134134e-05,
+      "loss": 4.0508,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7685334353623614e-05,
+      "loss": 4.039,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7676948406113094e-05,
+      "loss": 4.0535,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7668562458602574e-05,
+      "loss": 4.0428,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7660176511092054e-05,
+      "loss": 4.0579,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7651790563581534e-05,
+      "loss": 4.0632,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7643404616071014e-05,
+      "loss": 4.0462,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763503504736422e-05,
+      "loss": 4.0468,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762666547865744e-05,
+      "loss": 4.0504,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761827953114692e-05,
+      "loss": 4.0473,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76098935836364e-05,
+      "loss": 4.044,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760150763612588e-05,
+      "loss": 4.0583,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759312168861536e-05,
+      "loss": 4.0546,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758473574110484e-05,
+      "loss": 4.0551,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757634979359432e-05,
+      "loss": 4.0515,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75679638460838e-05,
+      "loss": 4.0556,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755957789857328e-05,
+      "loss": 4.0436,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755120832986649e-05,
+      "loss": 4.0429,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754282238235597e-05,
+      "loss": 4.0409,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753443643484545e-05,
+      "loss": 4.0499,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752605048733493e-05,
+      "loss": 4.0465,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751769729743187e-05,
+      "loss": 4.0307,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.750931134992135e-05,
+      "loss": 4.05,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.074104309082031,
+      "eval_runtime": 536.6879,
+      "eval_samples_per_second": 711.011,
+      "eval_steps_per_second": 22.22,
+      "step": 763190
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750092540241083e-05,
+      "loss": 4.0511,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749253945490031e-05,
+      "loss": 4.0528,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748415350738979e-05,
+      "loss": 4.0565,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.747576755987927e-05,
+      "loss": 4.0366,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.746738161236875e-05,
+      "loss": 4.0534,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745899566485823e-05,
+      "loss": 4.0437,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745062609615144e-05,
+      "loss": 4.0402,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744224014864092e-05,
+      "loss": 4.0395,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74338542011304e-05,
+      "loss": 4.0382,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742548463242361e-05,
+      "loss": 4.0498,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.741709868491309e-05,
+      "loss": 4.0496,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740871273740258e-05,
+      "loss": 4.041,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7400343168695786e-05,
+      "loss": 4.0496,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7391957221185266e-05,
+      "loss": 4.0323,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7383571273674746e-05,
+      "loss": 4.053,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7375185326164226e-05,
+      "loss": 4.0312,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7366799378653706e-05,
+      "loss": 4.038,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7358413431143186e-05,
+      "loss": 4.0438,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7350027483632666e-05,
+      "loss": 4.0433,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734164153612214e-05,
+      "loss": 4.0499,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7333271967415355e-05,
+      "loss": 4.0354,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7324886019904835e-05,
+      "loss": 4.0628,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7316500072394315e-05,
+      "loss": 4.0413,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7308114124883794e-05,
+      "loss": 4.0489,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7299728177373274e-05,
+      "loss": 4.039,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7291342229862754e-05,
+      "loss": 4.0504,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7282956282352234e-05,
+      "loss": 4.0454,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7274586713645443e-05,
+      "loss": 4.0375,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726620076613492e-05,
+      "loss": 4.0467,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.72578148186244e-05,
+      "loss": 4.0442,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724942887111388e-05,
+      "loss": 4.0272,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724104292360336e-05,
+      "loss": 4.0479,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.723265697609284e-05,
+      "loss": 4.049,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722427102858232e-05,
+      "loss": 4.0459,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.72158850810718e-05,
+      "loss": 4.0335,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720751551236501e-05,
+      "loss": 4.0539,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719914594365823e-05,
+      "loss": 4.04,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719075999614771e-05,
+      "loss": 4.0496,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718237404863719e-05,
+      "loss": 4.0419,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717398810112667e-05,
+      "loss": 4.0336,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716560215361615e-05,
+      "loss": 4.0356,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715721620610563e-05,
+      "loss": 4.0259,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714883025859511e-05,
+      "loss": 4.0443,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714044431108459e-05,
+      "loss": 4.0391,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71320747423778e-05,
+      "loss": 4.0324,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712368879486728e-05,
+      "loss": 4.0448,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7115319226160486e-05,
+      "loss": 4.0315,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7106933278649966e-05,
+      "loss": 4.0476,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709854733113945e-05,
+      "loss": 4.0423,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709016138362893e-05,
+      "loss": 4.0379,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708177543611841e-05,
+      "loss": 4.0387,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707338948860789e-05,
+      "loss": 4.0348,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706500354109737e-05,
+      "loss": 4.0416,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705663397239058e-05,
+      "loss": 4.027,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704824802488006e-05,
+      "loss": 4.0415,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703986207736954e-05,
+      "loss": 4.0342,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703147612985902e-05,
+      "loss": 4.0361,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70230901823485e-05,
+      "loss": 4.0324,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701472061364171e-05,
+      "loss": 4.0445,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700633466613119e-05,
+      "loss": 4.0296,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6997965097424406e-05,
+      "loss": 4.0267,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6989579149913886e-05,
+      "loss": 4.0345,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6981193202403366e-05,
+      "loss": 4.0284,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6972807254892846e-05,
+      "loss": 4.04,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6964421307382326e-05,
+      "loss": 4.0384,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69560353598718e-05,
+      "loss": 4.0379,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694764941236128e-05,
+      "loss": 4.0163,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693926346485076e-05,
+      "loss": 4.0379,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693087751734024e-05,
+      "loss": 4.0286,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692249156982972e-05,
+      "loss": 4.028,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69141056223192e-05,
+      "loss": 4.0305,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690571967480868e-05,
+      "loss": 4.0378,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689733372729816e-05,
+      "loss": 4.0306,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6888980537395104e-05,
+      "loss": 4.0493,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6880594589884584e-05,
+      "loss": 4.0364,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68722250211778e-05,
+      "loss": 4.0395,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686383907366727e-05,
+      "loss": 4.0377,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685545312615675e-05,
+      "loss": 4.0261,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.684706717864623e-05,
+      "loss": 4.0293,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683868123113571e-05,
+      "loss": 4.0364,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683029528362519e-05,
+      "loss": 4.0366,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682190933611467e-05,
+      "loss": 4.0432,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681352338860415e-05,
+      "loss": 4.0152,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680513744109363e-05,
+      "loss": 4.0197,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679675149358311e-05,
+      "loss": 4.0238,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678836554607259e-05,
+      "loss": 4.0383,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677999597736581e-05,
+      "loss": 4.0316,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677161002985529e-05,
+      "loss": 4.0401,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676322408234477e-05,
+      "loss": 4.0323,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675483813483425e-05,
+      "loss": 4.0335,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674645218732373e-05,
+      "loss": 4.0221,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673806623981321e-05,
+      "loss": 4.0302,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672968029230269e-05,
+      "loss": 4.0219,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672129434479216e-05,
+      "loss": 4.0306,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671292477608538e-05,
+      "loss": 4.0276,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670453882857486e-05,
+      "loss": 4.0307,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669615288106434e-05,
+      "loss": 4.0317,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668776693355381e-05,
+      "loss": 4.0408,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667938098604329e-05,
+      "loss": 4.0324,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6671011417336506e-05,
+      "loss": 4.0379,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6662625469825986e-05,
+      "loss": 4.0336,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6654239522315466e-05,
+      "loss": 4.0356,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664586995360868e-05,
+      "loss": 4.0306,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663748400609816e-05,
+      "loss": 4.037,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662911443739137e-05,
+      "loss": 4.0325,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662072848988085e-05,
+      "loss": 4.0265,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661234254237033e-05,
+      "loss": 4.0242,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660395659485981e-05,
+      "loss": 4.0178,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6595570647349284e-05,
+      "loss": 4.0307,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6587184699838764e-05,
+      "loss": 4.0304,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6578798752328244e-05,
+      "loss": 4.0367,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657041280481773e-05,
+      "loss": 4.0094,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.656204323611094e-05,
+      "loss": 4.0162,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.655365728860042e-05,
+      "loss": 4.034,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65452713410899e-05,
+      "loss": 4.0364,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653688539357938e-05,
+      "loss": 4.0432,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652849944606886e-05,
+      "loss": 4.0331,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652011349855834e-05,
+      "loss": 4.0223,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651172755104782e-05,
+      "loss": 4.0334,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65033416035373e-05,
+      "loss": 4.0309,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649497203483051e-05,
+      "loss": 4.0247,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648658608731999e-05,
+      "loss": 4.0243,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647820013980947e-05,
+      "loss": 4.0292,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646981419229895e-05,
+      "loss": 4.0256,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6461444623592164e-05,
+      "loss": 4.0292,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6453058676081644e-05,
+      "loss": 4.0272,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6444672728571124e-05,
+      "loss": 4.0344,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6436286781060604e-05,
+      "loss": 4.0209,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642791721235381e-05,
+      "loss": 4.0291,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641953126484329e-05,
+      "loss": 4.0267,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641114531733277e-05,
+      "loss": 4.0401,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640275936982225e-05,
+      "loss": 4.0408,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639438980111546e-05,
+      "loss": 4.0266,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638600385360494e-05,
+      "loss": 4.0273,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637761790609442e-05,
+      "loss": 4.0319,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63692319585839e-05,
+      "loss": 4.0288,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636086238987712e-05,
+      "loss": 4.0239,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63524764423666e-05,
+      "loss": 4.0389,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634409049485608e-05,
+      "loss": 4.0358,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.633570454734556e-05,
+      "loss": 4.0373,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632731859983504e-05,
+      "loss": 4.0318,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631894903112825e-05,
+      "loss": 4.036,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631056308361773e-05,
+      "loss": 4.0197,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630217713610721e-05,
+      "loss": 4.0249,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629379118859669e-05,
+      "loss": 4.0251,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628540524108617e-05,
+      "loss": 4.0283,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627701929357564e-05,
+      "loss": 4.0259,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6268649724868856e-05,
+      "loss": 4.0148,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626028015616207e-05,
+      "loss": 4.0306,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.062036037445068,
+      "eval_runtime": 525.7747,
+      "eval_samples_per_second": 725.769,
+      "eval_steps_per_second": 22.681,
+      "step": 839509
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625189420865155e-05,
+      "loss": 4.0392,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624350826114103e-05,
+      "loss": 4.0334,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623512231363051e-05,
+      "loss": 4.0342,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622673636611999e-05,
+      "loss": 4.02,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621835041860947e-05,
+      "loss": 4.0298,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6209964471098945e-05,
+      "loss": 4.0243,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6201578523588425e-05,
+      "loss": 4.0169,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6193192576077905e-05,
+      "loss": 4.0226,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6184806628567384e-05,
+      "loss": 4.0242,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6176420681056864e-05,
+      "loss": 4.0271,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6168034733546344e-05,
+      "loss": 4.0271,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6159648786035824e-05,
+      "loss": 4.0219,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6151262838525304e-05,
+      "loss": 4.0296,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6142876891014784e-05,
+      "loss": 4.0205,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613449094350427e-05,
+      "loss": 4.0289,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612610499599375e-05,
+      "loss": 4.015,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611771904848323e-05,
+      "loss": 4.0169,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610933310097271e-05,
+      "loss": 4.0254,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6100947153462184e-05,
+      "loss": 4.0216,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.60925775847554e-05,
+      "loss": 4.0282,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608419163724488e-05,
+      "loss": 4.0194,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607580568973436e-05,
+      "loss": 4.0433,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606741974222383e-05,
+      "loss": 4.0215,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605903379471331e-05,
+      "loss": 4.0268,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605064784720279e-05,
+      "loss": 4.023,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604226189969227e-05,
+      "loss": 4.0331,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.603387595218175e-05,
+      "loss": 4.0234,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602549000467124e-05,
+      "loss": 4.0185,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601712043596445e-05,
+      "loss": 4.0263,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.600873448845393e-05,
+      "loss": 4.0277,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600036491974714e-05,
+      "loss": 4.0109,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.599197897223662e-05,
+      "loss": 4.0249,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59835930247261e-05,
+      "loss": 4.0322,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597520707721558e-05,
+      "loss": 4.0297,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596682112970506e-05,
+      "loss": 4.0116,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595845156099827e-05,
+      "loss": 4.0293,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5950081992291476e-05,
+      "loss": 4.0257,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594169604478096e-05,
+      "loss": 4.0285,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593331009727044e-05,
+      "loss": 4.0208,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592492414975992e-05,
+      "loss": 4.0175,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59165382022494e-05,
+      "loss": 4.0159,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590815225473888e-05,
+      "loss": 4.0077,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589978268603209e-05,
+      "loss": 4.0237,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589139673852157e-05,
+      "loss": 4.0224,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588301079101105e-05,
+      "loss": 4.0159,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587462484350053e-05,
+      "loss": 4.023,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.586623889599001e-05,
+      "loss": 4.0128,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585785294847949e-05,
+      "loss": 4.0322,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584946700096897e-05,
+      "loss": 4.0227,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584108105345845e-05,
+      "loss": 4.0129,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583269510594793e-05,
+      "loss": 4.0172,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582430915843741e-05,
+      "loss": 4.0149,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581592321092689e-05,
+      "loss": 4.027,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580753726341637e-05,
+      "loss": 4.0035,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579916769470959e-05,
+      "loss": 4.026,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5790798126002796e-05,
+      "loss": 4.0147,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5782412178492276e-05,
+      "loss": 4.0172,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5774026230981756e-05,
+      "loss": 4.016,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5765640283471236e-05,
+      "loss": 4.025,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5757254335960716e-05,
+      "loss": 4.0085,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574886838845019e-05,
+      "loss": 4.0097,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5740498819743405e-05,
+      "loss": 4.0163,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5732112872232885e-05,
+      "loss": 4.0086,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5723726924722365e-05,
+      "loss": 4.0193,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5715340977211845e-05,
+      "loss": 4.0192,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5706955029701325e-05,
+      "loss": 4.0247,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5698569082190804e-05,
+      "loss": 3.9942,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5690183134680284e-05,
+      "loss": 4.018,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5681797187169764e-05,
+      "loss": 4.0129,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5673427618462974e-05,
+      "loss": 4.0064,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5665041670952453e-05,
+      "loss": 4.0115,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5656655723441933e-05,
+      "loss": 4.0225,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564826977593141e-05,
+      "loss": 4.0103,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563990020722462e-05,
+      "loss": 4.0318,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56315142597141e-05,
+      "loss": 4.0191,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562312831220358e-05,
+      "loss": 4.0214,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56147587434968e-05,
+      "loss": 4.0182,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560637279598628e-05,
+      "loss": 4.0077,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559798684847576e-05,
+      "loss": 4.0107,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558960090096524e-05,
+      "loss": 4.0181,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558121495345472e-05,
+      "loss": 4.0222,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55728290059442e-05,
+      "loss": 4.0195,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556445943723741e-05,
+      "loss": 4.0014,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555607348972689e-05,
+      "loss": 4.0046,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554768754221637e-05,
+      "loss": 4.0031,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553930159470585e-05,
+      "loss": 4.0219,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553091564719533e-05,
+      "loss": 4.0108,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552252969968481e-05,
+      "loss": 4.0222,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5514160130978016e-05,
+      "loss": 4.0144,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55057741834675e-05,
+      "loss": 4.0122,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549738823595698e-05,
+      "loss": 4.0056,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548900228844646e-05,
+      "loss": 4.0111,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548061634093594e-05,
+      "loss": 4.008,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547223039342542e-05,
+      "loss": 4.0074,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54638444459149e-05,
+      "loss": 4.0069,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545545849840438e-05,
+      "loss": 4.0166,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.544708892969759e-05,
+      "loss": 4.0149,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543870298218707e-05,
+      "loss": 4.0201,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543033341348028e-05,
+      "loss": 4.0161,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542194746596976e-05,
+      "loss": 4.0187,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.541356151845924e-05,
+      "loss": 4.0148,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.540517557094872e-05,
+      "loss": 4.0186,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53967896234382e-05,
+      "loss": 4.0116,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538840367592768e-05,
+      "loss": 4.0228,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538001772841716e-05,
+      "loss": 4.0116,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537163178090664e-05,
+      "loss": 4.0123,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5363262212199856e-05,
+      "loss": 4.0085,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.535487626468933e-05,
+      "loss": 3.9959,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5346506695982545e-05,
+      "loss": 4.0131,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5338120748472025e-05,
+      "loss": 4.0166,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53297348009615e-05,
+      "loss": 4.0131,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532134885345098e-05,
+      "loss": 3.995,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531296290594046e-05,
+      "loss": 3.997,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5304593337233674e-05,
+      "loss": 4.0175,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5296207389723154e-05,
+      "loss": 4.0171,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5287821442212634e-05,
+      "loss": 4.0233,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5279435494702114e-05,
+      "loss": 4.0169,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5271049547191594e-05,
+      "loss": 4.0103,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5262663599681074e-05,
+      "loss": 4.0118,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525429403097428e-05,
+      "loss": 4.0141,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524590808346376e-05,
+      "loss": 4.0058,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523752213595324e-05,
+      "loss": 4.004,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522913618844272e-05,
+      "loss": 4.0167,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52207502409322e-05,
+      "loss": 4.002,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521236429342168e-05,
+      "loss": 4.0156,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.520397834591116e-05,
+      "loss": 4.0057,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.519559239840064e-05,
+      "loss": 4.0167,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518722282969386e-05,
+      "loss": 4.0013,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517883688218334e-05,
+      "loss": 4.0144,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517045093467282e-05,
+      "loss": 4.009,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516208136596603e-05,
+      "loss": 4.0198,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515369541845551e-05,
+      "loss": 4.0215,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.514530947094499e-05,
+      "loss": 4.015,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513692352343447e-05,
+      "loss": 4.0078,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5128553954727677e-05,
+      "loss": 4.0127,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5120168007217157e-05,
+      "loss": 4.0112,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5111782059706636e-05,
+      "loss": 4.0079,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5103396112196116e-05,
+      "loss": 4.0203,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5095010164685596e-05,
+      "loss": 4.0173,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5086624217175076e-05,
+      "loss": 4.0203,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5078238269664556e-05,
+      "loss": 4.0164,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506985232215404e-05,
+      "loss": 4.0166,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5061466374643516e-05,
+      "loss": 4.0026,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505309680593673e-05,
+      "loss": 4.0061,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504471085842621e-05,
+      "loss": 4.0083,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5036324910915685e-05,
+      "loss": 4.0122,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50279553422089e-05,
+      "loss": 4.0101,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501958577350211e-05,
+      "loss": 3.9956,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501119982599159e-05,
+      "loss": 4.0118,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.051547527313232,
+      "eval_runtime": 551.4176,
+      "eval_samples_per_second": 692.018,
+      "eval_steps_per_second": 21.626,
+      "step": 915828
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500281387848107e-05,
+      "loss": 4.0191,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499442793097055e-05,
+      "loss": 4.0172,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498604198346003e-05,
+      "loss": 4.0146,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497765603594951e-05,
+      "loss": 4.0047,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496927008843899e-05,
+      "loss": 4.0134,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496088414092847e-05,
+      "loss": 4.0108,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495249819341795e-05,
+      "loss": 4.0013,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494411224590743e-05,
+      "loss": 4.0051,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493572629839691e-05,
+      "loss": 4.0039,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492734035088639e-05,
+      "loss": 4.008,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491895440337587e-05,
+      "loss": 4.0126,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491056845586535e-05,
+      "loss": 4.0081,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490218250835483e-05,
+      "loss": 4.0106,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489379656084431e-05,
+      "loss": 4.0019,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488541061333379e-05,
+      "loss": 4.0099,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487702466582327e-05,
+      "loss": 3.9991,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486863871831275e-05,
+      "loss": 4.0052,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4860269149605965e-05,
+      "loss": 4.0037,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4851883202095445e-05,
+      "loss": 4.0029,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4843497254584925e-05,
+      "loss": 4.0152,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4835111307074405e-05,
+      "loss": 4.0002,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482672535956388e-05,
+      "loss": 4.0213,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4818355790857094e-05,
+      "loss": 4.0063,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4809969843346574e-05,
+      "loss": 4.0105,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480158389583605e-05,
+      "loss": 4.0005,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479319794832553e-05,
+      "loss": 4.0168,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478481200081501e-05,
+      "loss": 4.0097,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477642605330449e-05,
+      "loss": 4.002,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476804010579397e-05,
+      "loss": 4.0058,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475965415828345e-05,
+      "loss": 4.0121,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475128458957666e-05,
+      "loss": 3.9941,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474291502086988e-05,
+      "loss": 4.0048,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473452907335935e-05,
+      "loss": 4.0154,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472614312584883e-05,
+      "loss": 4.0121,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471775717833831e-05,
+      "loss": 3.9946,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470938760963152e-05,
+      "loss": 4.0091,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470101804092474e-05,
+      "loss": 4.0114,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469263209341422e-05,
+      "loss": 4.0122,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46842461459037e-05,
+      "loss": 4.0019,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467586019839317e-05,
+      "loss": 4.005,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466747425088266e-05,
+      "loss": 4.0008,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465908830337214e-05,
+      "loss": 3.986,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465070235586162e-05,
+      "loss": 4.0107,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4642316408351097e-05,
+      "loss": 4.0047,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4633930460840577e-05,
+      "loss": 3.9963,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4625544513330056e-05,
+      "loss": 4.0086,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4617174944623266e-05,
+      "loss": 3.9959,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4608788997112746e-05,
+      "loss": 4.0144,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4600403049602225e-05,
+      "loss": 4.0082,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4592017102091705e-05,
+      "loss": 3.9926,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4583631154581185e-05,
+      "loss": 4.0027,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575245207070665e-05,
+      "loss": 3.9983,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4566859259560145e-05,
+      "loss": 4.01,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4558473312049625e-05,
+      "loss": 3.9873,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550087364539105e-05,
+      "loss": 4.0102,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454171779583232e-05,
+      "loss": 3.9927,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45333318483218e-05,
+      "loss": 4.0011,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452494590081128e-05,
+      "loss": 4.0015,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451655995330076e-05,
+      "loss": 4.0102,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450819038459397e-05,
+      "loss": 3.996,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449980443708345e-05,
+      "loss": 3.9884,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449141848957293e-05,
+      "loss": 3.9996,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448303254206241e-05,
+      "loss": 3.9945,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447464659455188e-05,
+      "loss": 4.0006,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446629340464883e-05,
+      "loss": 4.006,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445790745713831e-05,
+      "loss": 4.0047,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444952150962779e-05,
+      "loss": 3.9845,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4441135562117275e-05,
+      "loss": 3.9992,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4432749614606755e-05,
+      "loss": 3.998,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4424380045899964e-05,
+      "loss": 3.9917,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4415994098389444e-05,
+      "loss": 3.9977,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4407608150878924e-05,
+      "loss": 4.0028,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439923858217213e-05,
+      "loss": 3.9975,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439085263466161e-05,
+      "loss": 4.0058,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438246668715109e-05,
+      "loss": 4.0062,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437408073964057e-05,
+      "loss": 4.0073,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436569479213005e-05,
+      "loss": 4.0057,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435730884461953e-05,
+      "loss": 3.9899,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434892289710901e-05,
+      "loss": 3.9932,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434053694959849e-05,
+      "loss": 4.0021,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433216738089171e-05,
+      "loss": 4.0066,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432378143338118e-05,
+      "loss": 4.0021,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431539548587066e-05,
+      "loss": 3.9809,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430700953836014e-05,
+      "loss": 3.9935,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429862359084962e-05,
+      "loss": 3.9857,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429025402214283e-05,
+      "loss": 4.0027,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428186807463231e-05,
+      "loss": 3.9972,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427348212712179e-05,
+      "loss": 4.0095,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4265112558415006e-05,
+      "loss": 3.996,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425672661090448e-05,
+      "loss": 3.9967,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4248340663393966e-05,
+      "loss": 3.9883,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4239954715883446e-05,
+      "loss": 3.9946,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4231568768372926e-05,
+      "loss": 3.9944,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4223182820862406e-05,
+      "loss": 3.9935,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4214796873351886e-05,
+      "loss": 3.9883,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4206410925841366e-05,
+      "loss": 3.9982,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4198024978330846e-05,
+      "loss": 4.0,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4189639030820326e-05,
+      "loss": 4.0061,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4181253083309806e-05,
+      "loss": 3.9998,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4172883514603015e-05,
+      "loss": 4.0052,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4164497567092495e-05,
+      "loss": 3.9997,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156111619581975e-05,
+      "loss": 3.9989,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147725672071455e-05,
+      "loss": 3.9983,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139356103364664e-05,
+      "loss": 4.0071,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4130970155854144e-05,
+      "loss": 3.9952,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.412260058714736e-05,
+      "loss": 4.0009,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411421463963684e-05,
+      "loss": 3.9869,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410582869212632e-05,
+      "loss": 3.9839,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40974427446158e-05,
+      "loss": 3.9948,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408905679710528e-05,
+      "loss": 4.0012,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408067084959476e-05,
+      "loss": 3.9977,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407228490208424e-05,
+      "loss": 3.9796,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406391533337745e-05,
+      "loss": 3.9796,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405552938586693e-05,
+      "loss": 4.0068,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404715981716014e-05,
+      "loss": 4.0041,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403877386964962e-05,
+      "loss": 4.0046,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40303879221391e-05,
+      "loss": 4.0005,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4022001974628584e-05,
+      "loss": 3.9974,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4013616027118064e-05,
+      "loss": 3.9973,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4005230079607544e-05,
+      "loss": 3.9978,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399684413209702e-05,
+      "loss": 3.9868,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39884581845865e-05,
+      "loss": 3.9898,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398007223707598e-05,
+      "loss": 4.0024,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397170266836919e-05,
+      "loss": 3.9842,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3963316720858666e-05,
+      "loss": 3.9971,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3954930773348146e-05,
+      "loss": 3.9944,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3946544825837626e-05,
+      "loss": 3.9982,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3938175257130835e-05,
+      "loss": 3.9882,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392980568842405e-05,
+      "loss": 3.9984,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392141974091354e-05,
+      "loss": 3.9955,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391303379340302e-05,
+      "loss": 3.9991,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.390464784589249e-05,
+      "loss": 4.0054,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.389626189838197e-05,
+      "loss": 4.0022,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388787595087145e-05,
+      "loss": 3.995,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387949000336093e-05,
+      "loss": 3.996,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387110405585041e-05,
+      "loss": 3.9963,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386271810833989e-05,
+      "loss": 3.9872,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385433216082937e-05,
+      "loss": 4.0065,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384596259212258e-05,
+      "loss": 4.0031,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383757664461206e-05,
+      "loss": 4.0062,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382919069710154e-05,
+      "loss": 3.9999,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382080474959102e-05,
+      "loss": 4.0033,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3812418802080507e-05,
+      "loss": 3.9847,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3804032854569986e-05,
+      "loss": 3.9938,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3795646907059466e-05,
+      "loss": 3.9906,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3787260959548946e-05,
+      "loss": 3.9992,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3778907769645885e-05,
+      "loss": 3.9951,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3770521822135365e-05,
+      "loss": 3.9785,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3762135874624845e-05,
+      "loss": 3.992,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.042651176452637,
+      "eval_runtime": 539.7087,
+      "eval_samples_per_second": 707.031,
+      "eval_steps_per_second": 22.095,
+      "step": 992147
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3753749927114324e-05,
+      "loss": 3.999,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3745363979603804e-05,
+      "loss": 3.9995,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3736994410897014e-05,
+      "loss": 4.0041,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3728608463386493e-05,
+      "loss": 3.9884,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3720222515875973e-05,
+      "loss": 3.9993,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371183656836546e-05,
+      "loss": 3.9989,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370346699965867e-05,
+      "loss": 3.9781,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369509743095188e-05,
+      "loss": 3.9935,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368671148344136e-05,
+      "loss": 3.9919,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367832553593084e-05,
+      "loss": 3.9878,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366993958842032e-05,
+      "loss": 3.9988,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36615536409098e-05,
+      "loss": 3.9955,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365316769339928e-05,
+      "loss": 3.9864,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364478174588876e-05,
+      "loss": 3.9919,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363641217718197e-05,
+      "loss": 3.9956,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362802622967145e-05,
+      "loss": 3.9907,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361964028216093e-05,
+      "loss": 3.9836,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3611254334650414e-05,
+      "loss": 3.9887,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3602868387139894e-05,
+      "loss": 3.9886,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3594482439629374e-05,
+      "loss": 3.999,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3586096492118854e-05,
+      "loss": 3.9889,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357771054460833e-05,
+      "loss": 4.0022,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356932459709781e-05,
+      "loss": 3.9902,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356093864958729e-05,
+      "loss": 3.9978,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355255270207677e-05,
+      "loss": 3.9888,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3544183133369976e-05,
+      "loss": 3.9975,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3535797185859456e-05,
+      "loss": 3.9948,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3527411238348936e-05,
+      "loss": 3.9885,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3519025290838416e-05,
+      "loss": 3.9907,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3510639343327896e-05,
+      "loss": 3.9983,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3502253395817376e-05,
+      "loss": 3.9816,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349386744830686e-05,
+      "loss": 3.9899,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.348549787960007e-05,
+      "loss": 4.0011,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347711193208955e-05,
+      "loss": 3.9983,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346872598457903e-05,
+      "loss": 3.9793,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346034003706851e-05,
+      "loss": 3.9931,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345197046836172e-05,
+      "loss": 4.0003,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34435845208512e-05,
+      "loss": 3.9945,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343519857334068e-05,
+      "loss": 3.9849,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342681262583016e-05,
+      "loss": 3.9906,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341842667831964e-05,
+      "loss": 3.9866,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341004073080912e-05,
+      "loss": 3.9712,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34016547832986e-05,
+      "loss": 3.9954,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339326883578808e-05,
+      "loss": 3.9903,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3384899267081296e-05,
+      "loss": 3.9843,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3376513319570776e-05,
+      "loss": 3.9923,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3368127372060256e-05,
+      "loss": 3.9823,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3359741424549736e-05,
+      "loss": 3.9992,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3351371855842945e-05,
+      "loss": 3.993,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3343002287136154e-05,
+      "loss": 3.9791,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3334616339625634e-05,
+      "loss": 3.9894,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3326230392115114e-05,
+      "loss": 3.9834,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3317844444604594e-05,
+      "loss": 3.9938,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309458497094074e-05,
+      "loss": 3.9743,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330108892838728e-05,
+      "loss": 3.9977,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32927193596805e-05,
+      "loss": 3.9776,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.328433341216998e-05,
+      "loss": 3.9862,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.327594746465946e-05,
+      "loss": 3.9904,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326756151714894e-05,
+      "loss": 3.9887,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325917556963842e-05,
+      "loss": 3.9862,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32507896221279e-05,
+      "loss": 3.9736,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324240367461738e-05,
+      "loss": 3.9832,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323401772710686e-05,
+      "loss": 3.9838,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322563177959633e-05,
+      "loss": 3.9831,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321724583208581e-05,
+      "loss": 3.9928,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320885988457529e-05,
+      "loss": 3.9847,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320047393706477e-05,
+      "loss": 3.9735,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319210436835799e-05,
+      "loss": 3.9796,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318371842084747e-05,
+      "loss": 3.9883,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317533247333695e-05,
+      "loss": 3.973,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316694652582643e-05,
+      "loss": 3.9831,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315856057831591e-05,
+      "loss": 3.9938,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3150191009609116e-05,
+      "loss": 3.9809,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3141805062098596e-05,
+      "loss": 3.9895,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3133435493391805e-05,
+      "loss": 3.9935,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3125049545881285e-05,
+      "loss": 3.9925,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3116663598370765e-05,
+      "loss": 3.9922,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3108277650860245e-05,
+      "loss": 3.9753,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309990808215346e-05,
+      "loss": 3.9758,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309152213464294e-05,
+      "loss": 3.9855,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308313618713242e-05,
+      "loss": 3.9918,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30747502396219e-05,
+      "loss": 3.9878,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306636429211138e-05,
+      "loss": 3.9674,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305799472340459e-05,
+      "loss": 3.9802,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304960877589407e-05,
+      "loss": 3.9729,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304122282838355e-05,
+      "loss": 3.9837,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303283688087303e-05,
+      "loss": 3.9842,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302445093336251e-05,
+      "loss": 3.9968,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301608136465572e-05,
+      "loss": 3.9823,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30076954171452e-05,
+      "loss": 3.9834,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299930946963468e-05,
+      "loss": 3.9736,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299092352212416e-05,
+      "loss": 3.9773,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2982553953417375e-05,
+      "loss": 3.9788,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2974184384710584e-05,
+      "loss": 3.9833,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2965798437200064e-05,
+      "loss": 3.9714,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2957412489689544e-05,
+      "loss": 3.9833,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2949026542179024e-05,
+      "loss": 3.9845,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2940640594668504e-05,
+      "loss": 3.9903,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2932254647157984e-05,
+      "loss": 3.9893,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2923868699647464e-05,
+      "loss": 3.9901,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2915482752136944e-05,
+      "loss": 3.9855,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290711318343015e-05,
+      "loss": 3.9822,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289872723591963e-05,
+      "loss": 3.9821,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289035766721285e-05,
+      "loss": 3.9979,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288197171970233e-05,
+      "loss": 3.976,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287358577219181e-05,
+      "loss": 3.9883,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286519982468129e-05,
+      "loss": 3.9742,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285681387717077e-05,
+      "loss": 3.9671,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284842792966025e-05,
+      "loss": 3.9829,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284005836095346e-05,
+      "loss": 3.9869,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283167241344294e-05,
+      "loss": 3.984,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.282328646593242e-05,
+      "loss": 3.9625,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28149005184219e-05,
+      "loss": 3.9708,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280651457091138e-05,
+      "loss": 3.9824,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279812862340086e-05,
+      "loss": 3.9916,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278974267589034e-05,
+      "loss": 3.9913,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278135672837982e-05,
+      "loss": 3.9821,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27729707808693e-05,
+      "loss": 3.9853,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276458483335878e-05,
+      "loss": 3.9826,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275619888584826e-05,
+      "loss": 3.9806,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2747829317141466e-05,
+      "loss": 3.98,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2739443369630946e-05,
+      "loss": 3.9748,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2731057422120426e-05,
+      "loss": 3.9869,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2722671474609906e-05,
+      "loss": 3.9672,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2714301905903115e-05,
+      "loss": 3.984,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2705915958392595e-05,
+      "loss": 3.9789,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2697530010882075e-05,
+      "loss": 3.9877,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2689144063371555e-05,
+      "loss": 3.9703,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2680758115861035e-05,
+      "loss": 3.9838,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2672372168350515e-05,
+      "loss": 3.9811,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266398622084e-05,
+      "loss": 3.9847,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265560027332948e-05,
+      "loss": 3.9914,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264723070462269e-05,
+      "loss": 3.9906,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26388611359159e-05,
+      "loss": 3.975,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263047518840538e-05,
+      "loss": 3.9864,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262208924089486e-05,
+      "loss": 3.9808,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261370329338434e-05,
+      "loss": 3.9732,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260531734587382e-05,
+      "loss": 3.9934,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25969313983633e-05,
+      "loss": 3.987,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258854545085278e-05,
+      "loss": 3.9956,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258015950334226e-05,
+      "loss": 3.9818,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257178993463547e-05,
+      "loss": 3.9874,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2563403987124955e-05,
+      "loss": 3.9706,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2555018039614435e-05,
+      "loss": 3.9817,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2546632092103915e-05,
+      "loss": 3.9722,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2538262523397124e-05,
+      "loss": 3.9865,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2529876575886604e-05,
+      "loss": 3.981,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2521490628376084e-05,
+      "loss": 3.9646,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2513104680865564e-05,
+      "loss": 3.9801,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.03480339050293,
+      "eval_runtime": 539.8684,
+      "eval_samples_per_second": 706.822,
+      "eval_steps_per_second": 22.089,
+      "step": 1068466
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.250473511215877e-05,
+      "loss": 3.9889,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249634916464825e-05,
+      "loss": 3.9897,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248796321713773e-05,
+      "loss": 3.9851,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247957726962721e-05,
+      "loss": 3.9775,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247119132211669e-05,
+      "loss": 3.9851,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246280537460617e-05,
+      "loss": 3.9839,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245441942709565e-05,
+      "loss": 3.9668,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.244603347958513e-05,
+      "loss": 3.9763,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243764753207461e-05,
+      "loss": 3.9781,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242926158456409e-05,
+      "loss": 3.9762,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24208920158573e-05,
+      "loss": 3.9819,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.241250606834678e-05,
+      "loss": 3.9835,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240412012083626e-05,
+      "loss": 3.9717,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239573417332574e-05,
+      "loss": 3.979,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238734822581522e-05,
+      "loss": 3.9787,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23789622783047e-05,
+      "loss": 3.9742,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237057633079418e-05,
+      "loss": 3.9726,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236219038328366e-05,
+      "loss": 3.9759,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235380443577314e-05,
+      "loss": 3.9783,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234541848826262e-05,
+      "loss": 3.9763,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23370325407521e-05,
+      "loss": 3.9816,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232864659324159e-05,
+      "loss": 3.9842,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232026064573107e-05,
+      "loss": 3.9756,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231187469822054e-05,
+      "loss": 3.9825,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230348875071002e-05,
+      "loss": 3.9784,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22951028031995e-05,
+      "loss": 3.9868,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228673323449272e-05,
+      "loss": 3.9745,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227834728698219e-05,
+      "loss": 3.9806,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226996133947167e-05,
+      "loss": 3.9742,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226157539196115e-05,
+      "loss": 3.9856,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2253205823254366e-05,
+      "loss": 3.9675,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224481987574384e-05,
+      "loss": 3.9749,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2236433928233326e-05,
+      "loss": 3.9898,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222806435952654e-05,
+      "loss": 3.9804,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2219678412016015e-05,
+      "loss": 3.9665,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2211292464505495e-05,
+      "loss": 3.9799,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220292289579871e-05,
+      "loss": 3.9844,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219453694828819e-05,
+      "loss": 3.9826,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2186151000777664e-05,
+      "loss": 3.9734,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2177765053267144e-05,
+      "loss": 3.9763,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216939548456036e-05,
+      "loss": 3.9688,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216100953704984e-05,
+      "loss": 3.9601,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215262358953931e-05,
+      "loss": 3.9785,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214423764202879e-05,
+      "loss": 3.9765,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213586807332201e-05,
+      "loss": 3.9727,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212748212581149e-05,
+      "loss": 3.9738,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211909617830097e-05,
+      "loss": 3.9694,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211071023079045e-05,
+      "loss": 3.9842,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210232428327993e-05,
+      "loss": 3.982,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.209393833576941e-05,
+      "loss": 3.9668,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208555238825889e-05,
+      "loss": 3.9759,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207716644074837e-05,
+      "loss": 3.9679,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206879687204158e-05,
+      "loss": 3.9841,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206041092453106e-05,
+      "loss": 3.9619,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205202497702054e-05,
+      "loss": 3.9823,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204363902951002e-05,
+      "loss": 3.9614,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20352530819995e-05,
+      "loss": 3.9688,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202686713448898e-05,
+      "loss": 3.9819,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201849756578219e-05,
+      "loss": 3.9717,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201011161827167e-05,
+      "loss": 3.9742,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200172567076115e-05,
+      "loss": 3.9597,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199333972325063e-05,
+      "loss": 3.9686,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198495377574011e-05,
+      "loss": 3.9745,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197656782822959e-05,
+      "loss": 3.9667,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196818188071907e-05,
+      "loss": 3.9757,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195979593320855e-05,
+      "loss": 3.9724,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195142636450176e-05,
+      "loss": 3.959,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19430731745987e-05,
+      "loss": 3.9638,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193468722708819e-05,
+      "loss": 3.9805,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192630127957767e-05,
+      "loss": 3.9573,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191791533206715e-05,
+      "loss": 3.9683,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190952938455663e-05,
+      "loss": 3.9832,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190114343704611e-05,
+      "loss": 3.9621,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189275748953559e-05,
+      "loss": 3.9813,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1884371542025067e-05,
+      "loss": 3.9851,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1875985594514547e-05,
+      "loss": 3.9751,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1867599647004027e-05,
+      "loss": 3.9783,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18592136994935e-05,
+      "loss": 3.9602,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1850860509590445e-05,
+      "loss": 3.9684,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1842474562079925e-05,
+      "loss": 3.9686,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1834088614569405e-05,
+      "loss": 3.9778,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1825702667058885e-05,
+      "loss": 3.9719,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1817316719548365e-05,
+      "loss": 3.9581,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180893077203785e-05,
+      "loss": 3.9646,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1800544824527324e-05,
+      "loss": 3.9624,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1792158877016804e-05,
+      "loss": 3.966,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1783772929506284e-05,
+      "loss": 3.973,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1775386981995764e-05,
+      "loss": 3.9825,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1767001034485244e-05,
+      "loss": 3.9738,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1758615086974724e-05,
+      "loss": 3.9694,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175024551826793e-05,
+      "loss": 3.9604,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174185957075741e-05,
+      "loss": 3.9619,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173347362324689e-05,
+      "loss": 3.9682,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172508767573637e-05,
+      "loss": 3.9717,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171671810702959e-05,
+      "loss": 3.9599,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170833215951907e-05,
+      "loss": 3.968,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169994621200855e-05,
+      "loss": 3.9732,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169156026449803e-05,
+      "loss": 3.9766,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168317431698751e-05,
+      "loss": 3.9733,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167478836947699e-05,
+      "loss": 3.9769,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16664188007702e-05,
+      "loss": 3.977,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165803285325968e-05,
+      "loss": 3.9678,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164964690574916e-05,
+      "loss": 3.9689,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164126095823864e-05,
+      "loss": 3.9875,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163287501072812e-05,
+      "loss": 3.9646,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16244890632176e-05,
+      "loss": 3.9773,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161610311570708e-05,
+      "loss": 3.9626,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160773354700029e-05,
+      "loss": 3.9545,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1599347599489773e-05,
+      "loss": 3.9685,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1590961651979253e-05,
+      "loss": 3.9728,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1582575704468733e-05,
+      "loss": 3.9702,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157418975695821e-05,
+      "loss": 3.9498,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1565803809447687e-05,
+      "loss": 3.9619,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1557417861937166e-05,
+      "loss": 3.9718,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1549031914426646e-05,
+      "loss": 3.9764,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1540645966916126e-05,
+      "loss": 3.9739,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1532276398209335e-05,
+      "loss": 3.9711,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152390682950255e-05,
+      "loss": 3.9712,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151552088199203e-05,
+      "loss": 3.9718,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150713493448151e-05,
+      "loss": 3.9703,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149874898697099e-05,
+      "loss": 3.9678,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149036303946047e-05,
+      "loss": 3.959,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.148197709194995e-05,
+      "loss": 3.9753,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.147360752324316e-05,
+      "loss": 3.9555,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146522157573264e-05,
+      "loss": 3.9753,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145683562822212e-05,
+      "loss": 3.9647,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14484496807116e-05,
+      "loss": 3.9748,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144008011200481e-05,
+      "loss": 3.9549,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143169416449429e-05,
+      "loss": 3.9733,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142330821698377e-05,
+      "loss": 3.9702,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141492226947325e-05,
+      "loss": 3.9695,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140653632196273e-05,
+      "loss": 3.9806,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1398166753255945e-05,
+      "loss": 3.9728,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1389780805745425e-05,
+      "loss": 3.9649,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1381394858234905e-05,
+      "loss": 3.9774,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1373008910724385e-05,
+      "loss": 3.965,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1364622963213865e-05,
+      "loss": 3.9605,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1356237015703345e-05,
+      "loss": 3.9811,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1347851068192825e-05,
+      "loss": 3.9748,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1339465120682305e-05,
+      "loss": 3.9849,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1331095551975514e-05,
+      "loss": 3.9728,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1322709604464994e-05,
+      "loss": 3.9751,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13143400357582e-05,
+      "loss": 3.9597,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130597046705142e-05,
+      "loss": 3.9658,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.12975845195409e-05,
+      "loss": 3.9638,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128919857203038e-05,
+      "loss": 3.9714,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128081262451986e-05,
+      "loss": 3.9678,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127242667700934e-05,
+      "loss": 3.9566,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.126404072949882e-05,
+      "loss": 3.9659,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.027792453765869,
+      "eval_runtime": 562.0496,
+      "eval_samples_per_second": 678.928,
+      "eval_steps_per_second": 21.217,
+      "step": 1144785
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.12556547819883e-05,
+      "loss": 3.9658,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124726883447778e-05,
+      "loss": 3.981,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123888288696726e-05,
+      "loss": 3.9729,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123049693945674e-05,
+      "loss": 3.9651,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.122211099194622e-05,
+      "loss": 3.9751,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.12137250444357e-05,
+      "loss": 3.9702,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120535547572891e-05,
+      "loss": 3.9549,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119696952821839e-05,
+      "loss": 3.9625,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118858358070787e-05,
+      "loss": 3.9696,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118021401200108e-05,
+      "loss": 3.9622,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117184444329429e-05,
+      "loss": 3.97,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116345849578377e-05,
+      "loss": 3.9758,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115507254827325e-05,
+      "loss": 3.9552,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114668660076273e-05,
+      "loss": 3.9678,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113830065325221e-05,
+      "loss": 3.9673,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112991470574169e-05,
+      "loss": 3.9588,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1121528758231165e-05,
+      "loss": 3.9633,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1113142810720645e-05,
+      "loss": 3.9629,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1104756863210125e-05,
+      "loss": 3.9677,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1096370915699605e-05,
+      "loss": 3.9639,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1087984968189085e-05,
+      "loss": 3.968,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1079599020678565e-05,
+      "loss": 3.9738,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107122945197178e-05,
+      "loss": 3.9632,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.106284350446126e-05,
+      "loss": 3.9744,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105445755695074e-05,
+      "loss": 3.9666,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1046104367047686e-05,
+      "loss": 3.9736,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1037718419537166e-05,
+      "loss": 3.9611,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102933247202664e-05,
+      "loss": 3.9712,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102094652451612e-05,
+      "loss": 3.96,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.10125605770056e-05,
+      "loss": 3.9733,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.100417462949508e-05,
+      "loss": 3.9558,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099580506078829e-05,
+      "loss": 3.9621,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0987419113277774e-05,
+      "loss": 3.9776,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0979033165767254e-05,
+      "loss": 3.9705,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0970647218256734e-05,
+      "loss": 3.9563,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0962261270746214e-05,
+      "loss": 3.9651,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0953891702039423e-05,
+      "loss": 3.9699,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09455057545289e-05,
+      "loss": 3.9712,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093711980701838e-05,
+      "loss": 3.9659,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092873385950786e-05,
+      "loss": 3.966,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092034791199734e-05,
+      "loss": 3.9548,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091196196448682e-05,
+      "loss": 3.9562,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09035760169763e-05,
+      "loss": 3.9592,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089519006946578e-05,
+      "loss": 3.9662,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088680412195526e-05,
+      "loss": 3.9604,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087841817444474e-05,
+      "loss": 3.9612,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087006498454169e-05,
+      "loss": 3.9591,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086167903703117e-05,
+      "loss": 3.9651,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085329308952065e-05,
+      "loss": 3.9745,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084490714201013e-05,
+      "loss": 3.958,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083652119449961e-05,
+      "loss": 3.9586,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082813524698909e-05,
+      "loss": 3.9567,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081974929947857e-05,
+      "loss": 3.9707,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081136335196805e-05,
+      "loss": 3.9523,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.080297740445753e-05,
+      "loss": 3.9704,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0794591456947e-05,
+      "loss": 3.9514,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.078620550943648e-05,
+      "loss": 3.9583,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077781956192596e-05,
+      "loss": 3.9678,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076944999321918e-05,
+      "loss": 3.9646,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0761064045708657e-05,
+      "loss": 3.959,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0752678098198137e-05,
+      "loss": 3.9507,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0744292150687617e-05,
+      "loss": 3.9541,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073592258198083e-05,
+      "loss": 3.9641,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072755301327404e-05,
+      "loss": 3.9568,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071916706576352e-05,
+      "loss": 3.9568,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0710781118253e-05,
+      "loss": 3.9615,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0702395170742475e-05,
+      "loss": 3.954,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0694009223231955e-05,
+      "loss": 3.9494,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0685623275721434e-05,
+      "loss": 3.9673,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0677237328210914e-05,
+      "loss": 3.948,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0668851380700394e-05,
+      "loss": 3.951,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066046543318988e-05,
+      "loss": 3.9695,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065207948567936e-05,
+      "loss": 3.9525,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.064370991697257e-05,
+      "loss": 3.9671,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063532396946205e-05,
+      "loss": 3.9747,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062693802195153e-05,
+      "loss": 3.9609,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061855207444101e-05,
+      "loss": 3.967,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061019888453795e-05,
+      "loss": 3.9542,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060181293702743e-05,
+      "loss": 3.9521,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059342698951691e-05,
+      "loss": 3.9592,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058504104200639e-05,
+      "loss": 3.9676,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057665509449587e-05,
+      "loss": 3.9637,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056826914698535e-05,
+      "loss": 3.9455,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055988319947483e-05,
+      "loss": 3.9547,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0551497251964315e-05,
+      "loss": 3.9492,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0543111304453795e-05,
+      "loss": 3.9528,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0534741735747004e-05,
+      "loss": 3.9594,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0526355788236484e-05,
+      "loss": 3.9715,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051798621952969e-05,
+      "loss": 3.9618,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509600272019173e-05,
+      "loss": 3.9587,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0501214324508653e-05,
+      "loss": 3.9488,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0492828376998133e-05,
+      "loss": 3.9505,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0484442429487613e-05,
+      "loss": 3.9544,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0476056481977096e-05,
+      "loss": 3.9582,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0467670534466576e-05,
+      "loss": 3.9497,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0459284586956056e-05,
+      "loss": 3.953,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0450931397052994e-05,
+      "loss": 3.9628,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0442545449542474e-05,
+      "loss": 3.9639,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0434159502031957e-05,
+      "loss": 3.9622,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0425773554521437e-05,
+      "loss": 3.9589,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0417387607010917e-05,
+      "loss": 3.9693,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0409001659500397e-05,
+      "loss": 3.9552,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0400615711989877e-05,
+      "loss": 3.9578,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0392246143283086e-05,
+      "loss": 3.9734,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0383860195772566e-05,
+      "loss": 3.9562,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.037547424826205e-05,
+      "loss": 3.9655,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.036708830075153e-05,
+      "loss": 3.9524,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035870235324101e-05,
+      "loss": 3.942,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035033278453422e-05,
+      "loss": 3.9568,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.03419468370237e-05,
+      "loss": 3.9585,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.033356088951318e-05,
+      "loss": 3.9611,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.032517494200266e-05,
+      "loss": 3.9419,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0316788994492135e-05,
+      "loss": 3.9475,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0308403046981615e-05,
+      "loss": 3.9587,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0300017099471095e-05,
+      "loss": 3.9663,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0291631151960575e-05,
+      "loss": 3.9631,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0283245204450055e-05,
+      "loss": 3.9545,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0274875635743267e-05,
+      "loss": 3.9622,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0266489688232747e-05,
+      "loss": 3.9614,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0258103740722227e-05,
+      "loss": 3.9593,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0249717793211707e-05,
+      "loss": 3.9557,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0241331845701187e-05,
+      "loss": 3.9453,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0232945898190667e-05,
+      "loss": 3.9639,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.022459270828761e-05,
+      "loss": 3.942,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.021620676077709e-05,
+      "loss": 3.9644,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.020782081326657e-05,
+      "loss": 3.9527,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.019943486575605e-05,
+      "loss": 3.9618,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.019104891824553e-05,
+      "loss": 3.9419,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0182679349538738e-05,
+      "loss": 3.9636,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.017429340202822e-05,
+      "loss": 3.9576,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01659074545177e-05,
+      "loss": 3.9598,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.015752150700718e-05,
+      "loss": 3.9657,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.014913555949666e-05,
+      "loss": 3.9621,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.014076599078987e-05,
+      "loss": 3.9569,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.013238004327935e-05,
+      "loss": 3.9634,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.012399409576883e-05,
+      "loss": 3.9513,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115608148258313e-05,
+      "loss": 3.9547,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0107222200747793e-05,
+      "loss": 3.9665,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009886901084473e-05,
+      "loss": 3.9703,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009048306333421e-05,
+      "loss": 3.965,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.008209711582369e-05,
+      "loss": 3.9644,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073711168313175e-05,
+      "loss": 3.961,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0065325220802655e-05,
+      "loss": 3.9521,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0056939273292135e-05,
+      "loss": 3.9569,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0048553325781615e-05,
+      "loss": 3.9515,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0040167378271095e-05,
+      "loss": 3.9579,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0031781430760575e-05,
+      "loss": 3.9589,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0023395483250055e-05,
+      "loss": 3.9428,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0015009535739534e-05,
+      "loss": 3.9568,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.022145748138428,
+      "eval_runtime": 562.0389,
+      "eval_samples_per_second": 678.941,
+      "eval_steps_per_second": 21.217,
+      "step": 1221104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.0006639967032747e-05,
+      "loss": 3.9645,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998254019522227e-05,
+      "loss": 3.956,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989868072011707e-05,
+      "loss": 3.9526,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981482124501187e-05,
+      "loss": 3.9639,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9973096176990667e-05,
+      "loss": 3.9629,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964710229480147e-05,
+      "loss": 3.971,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956324281969623e-05,
+      "loss": 3.9549,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9947938334459103e-05,
+      "loss": 3.9511,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939552386948583e-05,
+      "loss": 3.9495,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931166439438063e-05,
+      "loss": 3.961,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922780491927543e-05,
+      "loss": 3.9573,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914394544417023e-05,
+      "loss": 3.9595,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9906008596906503e-05,
+      "loss": 3.9605,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9897639028199716e-05,
+      "loss": 3.9463,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889253080689195e-05,
+      "loss": 3.9507,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880867133178675e-05,
+      "loss": 3.9484,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9872481185668155e-05,
+      "loss": 3.9458,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9864095238157635e-05,
+      "loss": 3.9575,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9855709290647115e-05,
+      "loss": 3.9529,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9847323343136595e-05,
+      "loss": 3.9564,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9838937395626075e-05,
+      "loss": 3.9697,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830551448115555e-05,
+      "loss": 3.9534,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.982216550060504e-05,
+      "loss": 3.9616,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.981377955309451e-05,
+      "loss": 3.954,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.980539360558399e-05,
+      "loss": 3.9683,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9797024036877208e-05,
+      "loss": 3.9526,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788654468170417e-05,
+      "loss": 3.9573,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.97802685206599e-05,
+      "loss": 3.9479,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.977188257314938e-05,
+      "loss": 3.956,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.976349662563886e-05,
+      "loss": 3.95,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755110678128333e-05,
+      "loss": 3.9529,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9746724730617813e-05,
+      "loss": 3.9616,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973835516191103e-05,
+      "loss": 3.9599,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972996921440051e-05,
+      "loss": 3.9608,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9721583266889985e-05,
+      "loss": 3.9597,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.97132136981832e-05,
+      "loss": 3.9582,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.970482775067268e-05,
+      "loss": 3.9629,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9696441803162154e-05,
+      "loss": 3.9433,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9688055855651638e-05,
+      "loss": 3.954,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9679669908141118e-05,
+      "loss": 3.9384,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9671283960630598e-05,
+      "loss": 3.9496,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9662898013120078e-05,
+      "loss": 3.9468,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9654512065609558e-05,
+      "loss": 3.9641,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9646126118099038e-05,
+      "loss": 3.9516,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9637740170588517e-05,
+      "loss": 3.9477,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9629354223077997e-05,
+      "loss": 3.9542,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9620968275567477e-05,
+      "loss": 3.9568,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961259870686069e-05,
+      "loss": 3.9517,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.960421275935017e-05,
+      "loss": 3.9375,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.959584319064338e-05,
+      "loss": 3.9465,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.958745724313286e-05,
+      "loss": 3.967,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957907129562234e-05,
+      "loss": 3.9628,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9570685348111822e-05,
+      "loss": 3.9506,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562299400601302e-05,
+      "loss": 3.9513,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9553913453090782e-05,
+      "loss": 3.9587,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545527505580262e-05,
+      "loss": 3.9383,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537141558069742e-05,
+      "loss": 3.9635,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528755610559222e-05,
+      "loss": 3.9359,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.952038604185243e-05,
+      "loss": 3.9657,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9512000094341914e-05,
+      "loss": 3.9609,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503630525635123e-05,
+      "loss": 3.939,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9495244578124603e-05,
+      "loss": 3.9477,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486858630614083e-05,
+      "loss": 3.9593,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478472683103563e-05,
+      "loss": 3.9298,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9470086735593043e-05,
+      "loss": 3.943,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9461717166886256e-05,
+      "loss": 3.962,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9453331219375736e-05,
+      "loss": 3.9519,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9444945271865216e-05,
+      "loss": 3.9508,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9436559324354696e-05,
+      "loss": 3.9351,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942817337684417e-05,
+      "loss": 3.9461,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9419787429333652e-05,
+      "loss": 3.9395,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411401481823132e-05,
+      "loss": 3.9615,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9403015534312612e-05,
+      "loss": 3.9422,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.939464596560582e-05,
+      "loss": 3.9536,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.93862600180953e-05,
+      "loss": 3.9545,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377890449388517e-05,
+      "loss": 3.9587,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369504501877994e-05,
+      "loss": 3.947,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9361118554367474e-05,
+      "loss": 3.9487,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9352732606856954e-05,
+      "loss": 3.9557,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344346659346433e-05,
+      "loss": 3.9502,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9335977090639643e-05,
+      "loss": 3.9477,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.932760752193286e-05,
+      "loss": 3.9461,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931922157442234e-05,
+      "loss": 3.9546,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310835626911815e-05,
+      "loss": 3.9544,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302449679401295e-05,
+      "loss": 3.9375,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9294063731890775e-05,
+      "loss": 3.9464,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9285677784380255e-05,
+      "loss": 3.9458,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9277291836869735e-05,
+      "loss": 3.9531,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9268905889359215e-05,
+      "loss": 3.9457,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260536320652427e-05,
+      "loss": 3.951,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252150373141907e-05,
+      "loss": 3.9437,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243764425631387e-05,
+      "loss": 3.9382,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235378478120867e-05,
+      "loss": 3.9447,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9226992530610347e-05,
+      "loss": 3.9557,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921862296190356e-05,
+      "loss": 3.9498,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921023701439304e-05,
+      "loss": 3.9581,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.920185106688252e-05,
+      "loss": 3.9505,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9193465119372e-05,
+      "loss": 3.9424,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.918507917186148e-05,
+      "loss": 3.9461,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.917669322435096e-05,
+      "loss": 3.9574,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916830727684044e-05,
+      "loss": 3.9509,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915993770813365e-05,
+      "loss": 3.9584,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915155176062313e-05,
+      "loss": 3.9465,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.914316581311261e-05,
+      "loss": 3.9416,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.913477986560209e-05,
+      "loss": 3.9625,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.912639391809157e-05,
+      "loss": 3.9374,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.911800797058105e-05,
+      "loss": 3.9451,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910963840187426e-05,
+      "loss": 3.9502,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910125245436374e-05,
+      "loss": 3.947,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092866506853224e-05,
+      "loss": 3.9399,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9084480559342704e-05,
+      "loss": 3.9442,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076094611832177e-05,
+      "loss": 3.9409,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067708664321657e-05,
+      "loss": 3.9501,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059322716811137e-05,
+      "loss": 3.9465,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9050936769300617e-05,
+      "loss": 3.9461,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904256720059383e-05,
+      "loss": 3.9554,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.903418125308331e-05,
+      "loss": 3.9636,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.902579530557279e-05,
+      "loss": 3.9528,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9017425736866005e-05,
+      "loss": 3.9432,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900903978935548e-05,
+      "loss": 3.9424,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000653841844962e-05,
+      "loss": 3.9448,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.899226789433444e-05,
+      "loss": 3.9473,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898388194682392e-05,
+      "loss": 3.949,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897551237811713e-05,
+      "loss": 3.9439,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.896712643060661e-05,
+      "loss": 3.9523,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.895874048309609e-05,
+      "loss": 3.9458,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.895035453558557e-05,
+      "loss": 3.9492,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8941968588075054e-05,
+      "loss": 3.9443,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933599019368263e-05,
+      "loss": 3.944,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925213071857743e-05,
+      "loss": 3.9427,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916827124347223e-05,
+      "loss": 3.9569,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908441176836703e-05,
+      "loss": 3.9592,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8900055229326183e-05,
+      "loss": 3.9576,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891685660619395e-05,
+      "loss": 3.9397,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883299713108875e-05,
+      "loss": 3.9513,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8874913765598355e-05,
+      "loss": 3.9484,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8866527818087835e-05,
+      "loss": 3.9513,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8858141870577315e-05,
+      "loss": 3.946,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849755923066795e-05,
+      "loss": 3.9609,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841386354360008e-05,
+      "loss": 3.9444,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8833000406849488e-05,
+      "loss": 3.9639,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8824614459338968e-05,
+      "loss": 3.9444,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8816228511828447e-05,
+      "loss": 3.958,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807842564317927e-05,
+      "loss": 3.9467,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8799456616807407e-05,
+      "loss": 3.9403,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791070669296887e-05,
+      "loss": 3.9511,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.87827011005901e-05,
+      "loss": 3.9362,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.877431515307958e-05,
+      "loss": 3.9452,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.876594558437279e-05,
+      "loss": 3.9505,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.875755963686227e-05,
+      "loss": 3.9456,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.016632080078125,
+      "eval_runtime": 305.9882,
+      "eval_samples_per_second": 1247.077,
+      "eval_steps_per_second": 38.972,
+      "step": 1297424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874917368935175e-05,
+      "loss": 3.9474,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874078774184123e-05,
+      "loss": 3.946,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.873240179433071e-05,
+      "loss": 3.9524,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.872401584682019e-05,
+      "loss": 3.9486,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715629899309665e-05,
+      "loss": 3.9625,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707243951799145e-05,
+      "loss": 3.9434,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8698858004288625e-05,
+      "loss": 3.9438,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8690472056778105e-05,
+      "loss": 3.9396,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8682086109267585e-05,
+      "loss": 3.9502,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8673700161757065e-05,
+      "loss": 3.947,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.866531421424655e-05,
+      "loss": 3.9491,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8656944645539757e-05,
+      "loss": 3.9476,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8648575076832967e-05,
+      "loss": 3.9378,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640189129322446e-05,
+      "loss": 3.9481,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631803181811926e-05,
+      "loss": 3.9357,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.862341723430141e-05,
+      "loss": 3.9397,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.861503128679089e-05,
+      "loss": 3.9436,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.860664533928037e-05,
+      "loss": 3.9421,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.859825939176985e-05,
+      "loss": 3.9462,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858988982306306e-05,
+      "loss": 3.9644,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858150387555254e-05,
+      "loss": 3.9449,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.857311792804202e-05,
+      "loss": 3.951,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564731980531502e-05,
+      "loss": 3.9459,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.855636241182471e-05,
+      "loss": 3.9546,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.854797646431419e-05,
+      "loss": 3.9454,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.853959051680367e-05,
+      "loss": 3.9466,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.853120456929315e-05,
+      "loss": 3.9369,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.852281862178263e-05,
+      "loss": 3.945,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.851443267427211e-05,
+      "loss": 3.9362,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8506046726761594e-05,
+      "loss": 3.9471,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497660779251074e-05,
+      "loss": 3.9502,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489291210544283e-05,
+      "loss": 3.9475,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8480921641837492e-05,
+      "loss": 3.9541,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472535694326972e-05,
+      "loss": 3.9462,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464166125620185e-05,
+      "loss": 3.9516,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455780178109665e-05,
+      "loss": 3.9517,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447394230599145e-05,
+      "loss": 3.9392,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8439008283088625e-05,
+      "loss": 3.9408,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8430622335578105e-05,
+      "loss": 3.9279,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8422236388067585e-05,
+      "loss": 3.9419,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8413850440557065e-05,
+      "loss": 3.9401,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8405464493046548e-05,
+      "loss": 3.9536,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8397078545536028e-05,
+      "loss": 3.941,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.83886925980255e-05,
+      "loss": 3.9406,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.838030665051498e-05,
+      "loss": 3.9466,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.837192070300446e-05,
+      "loss": 3.9445,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8363551134297673e-05,
+      "loss": 3.941,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8355165186787153e-05,
+      "loss": 3.9287,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8346779239276633e-05,
+      "loss": 3.9336,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8338393291766113e-05,
+      "loss": 3.9575,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8330007344255593e-05,
+      "loss": 3.9528,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8321637775548802e-05,
+      "loss": 3.9374,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8313251828038286e-05,
+      "loss": 3.9409,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.83048822593315e-05,
+      "loss": 3.9524,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296496311820975e-05,
+      "loss": 3.93,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8288110364310455e-05,
+      "loss": 3.9501,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8279724416799935e-05,
+      "loss": 3.9266,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271338469289415e-05,
+      "loss": 3.9562,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8262952521778895e-05,
+      "loss": 3.947,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8254566574268378e-05,
+      "loss": 3.9376,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8246180626757858e-05,
+      "loss": 3.9309,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8237794679247338e-05,
+      "loss": 3.9512,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8229408731736818e-05,
+      "loss": 3.9212,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8221022784226298e-05,
+      "loss": 3.9338,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212636836715778e-05,
+      "loss": 3.9506,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8204267268008987e-05,
+      "loss": 3.9456,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.819588132049847e-05,
+      "loss": 3.9407,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.818751175179168e-05,
+      "loss": 3.9262,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817912580428116e-05,
+      "loss": 3.9354,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817073985677064e-05,
+      "loss": 3.9324,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.816235390926012e-05,
+      "loss": 3.947,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81539679617496e-05,
+      "loss": 3.9372,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.814558201423908e-05,
+      "loss": 3.9419,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.813719606672856e-05,
+      "loss": 3.9449,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128810119218036e-05,
+      "loss": 3.95,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.812044055051125e-05,
+      "loss": 3.9335,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.811205460300073e-05,
+      "loss": 3.9419,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.810366865549021e-05,
+      "loss": 3.9442,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8095282707979688e-05,
+      "loss": 3.9432,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8086913139272904e-05,
+      "loss": 3.9384,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8078527191762384e-05,
+      "loss": 3.9382,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070141244251857e-05,
+      "loss": 3.9447,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8061755296741337e-05,
+      "loss": 3.9423,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8053369349230817e-05,
+      "loss": 3.9301,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8044999780524033e-05,
+      "loss": 3.9332,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.803661383301351e-05,
+      "loss": 3.9394,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.802822788550299e-05,
+      "loss": 3.9443,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801984193799247e-05,
+      "loss": 3.9391,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801145599048195e-05,
+      "loss": 3.9374,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.800307004297143e-05,
+      "loss": 3.9331,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.799468409546091e-05,
+      "loss": 3.9299,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.798629814795039e-05,
+      "loss": 3.9307,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7977912200439872e-05,
+      "loss": 3.9431,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796954263173308e-05,
+      "loss": 3.9378,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796115668422256e-05,
+      "loss": 3.9465,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.795278711551577e-05,
+      "loss": 3.9453,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.794440116800525e-05,
+      "loss": 3.9336,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7936015220494734e-05,
+      "loss": 3.9346,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7927629272984214e-05,
+      "loss": 3.944,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919243325473694e-05,
+      "loss": 3.9463,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7910873756766903e-05,
+      "loss": 3.9438,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7902504188060112e-05,
+      "loss": 3.9395,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7894118240549595e-05,
+      "loss": 3.9319,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885732293039075e-05,
+      "loss": 3.9517,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877346345528555e-05,
+      "loss": 3.9261,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7868960398018035e-05,
+      "loss": 3.9386,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7860574450507515e-05,
+      "loss": 3.9365,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7852188502996995e-05,
+      "loss": 3.9382,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7843802555486475e-05,
+      "loss": 3.9297,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7835432986779687e-05,
+      "loss": 3.9386,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7827047039269167e-05,
+      "loss": 3.9306,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7818661091758647e-05,
+      "loss": 3.9397,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810275144248127e-05,
+      "loss": 3.9366,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7801889196737607e-05,
+      "loss": 3.9323,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7793503249227087e-05,
+      "loss": 3.9495,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7785117301716567e-05,
+      "loss": 3.953,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7776731354206047e-05,
+      "loss": 3.9424,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768345406695524e-05,
+      "loss": 3.9372,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7759959459185004e-05,
+      "loss": 3.9332,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.775158989047822e-05,
+      "loss": 3.9363,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7743203942967693e-05,
+      "loss": 3.9369,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734817995457173e-05,
+      "loss": 3.9441,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726432047946656e-05,
+      "loss": 3.9308,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7718062479239872e-05,
+      "loss": 3.9391,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7709676531729345e-05,
+      "loss": 3.9389,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7701290584218825e-05,
+      "loss": 3.9439,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7692904636708305e-05,
+      "loss": 3.9281,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.768453506800152e-05,
+      "loss": 3.9374,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7676149120490997e-05,
+      "loss": 3.9354,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667763172980477e-05,
+      "loss": 3.9438,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7659377225469957e-05,
+      "loss": 3.9486,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7650991277959437e-05,
+      "loss": 3.949,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7642621709252646e-05,
+      "loss": 3.9316,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634235761742126e-05,
+      "loss": 3.941,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.762584981423161e-05,
+      "loss": 3.9428,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.761746386672109e-05,
+      "loss": 3.9407,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.760907791921057e-05,
+      "loss": 3.9383,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.760070835050378e-05,
+      "loss": 3.9475,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.759232240299326e-05,
+      "loss": 3.9358,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.758393645548274e-05,
+      "loss": 3.9562,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.757555050797222e-05,
+      "loss": 3.931,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.75671645604617e-05,
+      "loss": 3.9485,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7558778612951182e-05,
+      "loss": 3.9383,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7550392665440662e-05,
+      "loss": 3.9326,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7542006717930142e-05,
+      "loss": 3.9398,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.753363714922335e-05,
+      "loss": 3.9305,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7525267580516563e-05,
+      "loss": 3.9346,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7516881633006043e-05,
+      "loss": 3.9392,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508495685495523e-05,
+      "loss": 3.9318,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0120720863342285,
+      "eval_runtime": 295.7178,
+      "eval_samples_per_second": 1290.389,
+      "eval_steps_per_second": 40.326,
+      "step": 1373744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7500109737985003e-05,
+      "loss": 3.9389,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491723790474483e-05,
+      "loss": 3.9328,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483337842963963e-05,
+      "loss": 3.9459,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7474951895453443e-05,
+      "loss": 3.9382,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466565947942923e-05,
+      "loss": 3.9518,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458180000432403e-05,
+      "loss": 3.9344,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744979405292188e-05,
+      "loss": 3.9368,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744140810541136e-05,
+      "loss": 3.9271,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743302215790084e-05,
+      "loss": 3.9429,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742463621039032e-05,
+      "loss": 3.9362,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.74162502628798e-05,
+      "loss": 3.9339,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.740786431536928e-05,
+      "loss": 3.9445,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7399494746662492e-05,
+      "loss": 3.9244,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7391108799151972e-05,
+      "loss": 3.9391,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.738272285164145e-05,
+      "loss": 3.9235,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.737433690413093e-05,
+      "loss": 3.931,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.736595095662041e-05,
+      "loss": 3.9318,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.735756500910989e-05,
+      "loss": 3.9345,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734917906159937e-05,
+      "loss": 3.9339,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734079311408885e-05,
+      "loss": 3.9552,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332423545382064e-05,
+      "loss": 3.9326,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7324037597871544e-05,
+      "loss": 3.9391,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7315651650361024e-05,
+      "loss": 3.9373,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7307265702850504e-05,
+      "loss": 3.9453,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7298879755339984e-05,
+      "loss": 3.9339,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290493807829464e-05,
+      "loss": 3.9426,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7282107860318944e-05,
+      "loss": 3.9242,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.727372191280842e-05,
+      "loss": 3.9357,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7265352344101636e-05,
+      "loss": 3.9249,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7256982775394845e-05,
+      "loss": 3.9413,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7248596827884325e-05,
+      "loss": 3.937,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7240210880373805e-05,
+      "loss": 3.9403,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.723182493286329e-05,
+      "loss": 3.9472,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223455364156498e-05,
+      "loss": 3.9353,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7215085795449707e-05,
+      "loss": 3.942,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7206699847939187e-05,
+      "loss": 3.9393,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7198313900428667e-05,
+      "loss": 3.9332,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718992795291815e-05,
+      "loss": 3.9265,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718154200540763e-05,
+      "loss": 3.9224,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.717315605789711e-05,
+      "loss": 3.9291,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.716477011038659e-05,
+      "loss": 3.9315,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.715638416287607e-05,
+      "loss": 3.9428,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7147998215365543e-05,
+      "loss": 3.936,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139612267855026e-05,
+      "loss": 3.9268,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131226320344506e-05,
+      "loss": 3.9394,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122840372833986e-05,
+      "loss": 3.9307,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.711448718293093e-05,
+      "loss": 3.9318,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.710610123542041e-05,
+      "loss": 3.9236,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709771528790989e-05,
+      "loss": 3.921,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089329340399368e-05,
+      "loss": 3.9482,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7080943392888848e-05,
+      "loss": 3.9416,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7072557445378328e-05,
+      "loss": 3.928,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7064171497867808e-05,
+      "loss": 3.9316,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7055785550357287e-05,
+      "loss": 3.9406,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7047415981650497e-05,
+      "loss": 3.9176,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703903003413998e-05,
+      "loss": 3.941,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703064408662946e-05,
+      "loss": 3.9177,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.702225813911894e-05,
+      "loss": 3.9448,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.701387219160842e-05,
+      "loss": 3.9413,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.70054862440979e-05,
+      "loss": 3.9265,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.699710029658738e-05,
+      "loss": 3.9211,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698871434907686e-05,
+      "loss": 3.9424,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698034478037007e-05,
+      "loss": 3.9099,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.697197521166328e-05,
+      "loss": 3.9272,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.696358926415276e-05,
+      "loss": 3.9402,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.695520331664224e-05,
+      "loss": 3.9374,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.694681736913172e-05,
+      "loss": 3.9293,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.69384314216212e-05,
+      "loss": 3.9171,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.693004547411068e-05,
+      "loss": 3.9238,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921675905403894e-05,
+      "loss": 3.9222,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913289957893373e-05,
+      "loss": 3.934,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6904904010382853e-05,
+      "loss": 3.9258,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6896518062872333e-05,
+      "loss": 3.9352,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6888132115361813e-05,
+      "loss": 3.9366,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6879762546655022e-05,
+      "loss": 3.94,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871376599144506e-05,
+      "loss": 3.9247,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6862990651633986e-05,
+      "loss": 3.9298,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6854604704123466e-05,
+      "loss": 3.9368,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6846218756612946e-05,
+      "loss": 3.9308,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6837832809102426e-05,
+      "loss": 3.9291,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6829463240395635e-05,
+      "loss": 3.9314,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6821077292885115e-05,
+      "loss": 3.9346,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6812691345374598e-05,
+      "loss": 3.9335,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804305397864078e-05,
+      "loss": 3.9186,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6795935829157287e-05,
+      "loss": 3.9259,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787549881646767e-05,
+      "loss": 3.9275,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779163934136247e-05,
+      "loss": 3.9346,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770777986625727e-05,
+      "loss": 3.9303,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.676240841791894e-05,
+      "loss": 3.9301,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.675402247040842e-05,
+      "loss": 3.9239,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.67456365228979e-05,
+      "loss": 3.9193,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737250575387372e-05,
+      "loss": 3.923,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728864627876852e-05,
+      "loss": 3.9321,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.672049505917007e-05,
+      "loss": 3.9377,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.671210911165955e-05,
+      "loss": 3.9378,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703723164149025e-05,
+      "loss": 3.9313,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695337216638505e-05,
+      "loss": 3.9268,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6686951269127985e-05,
+      "loss": 3.9261,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.66785817004212e-05,
+      "loss": 3.9307,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670195752910677e-05,
+      "loss": 3.9383,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6661809805400157e-05,
+      "loss": 3.9346,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653423857889637e-05,
+      "loss": 3.9296,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6645037910379117e-05,
+      "loss": 3.9228,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6636668341672326e-05,
+      "loss": 3.9396,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6628282394161806e-05,
+      "loss": 3.9183,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661989644665129e-05,
+      "loss": 3.9307,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661151049914077e-05,
+      "loss": 3.9277,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.660312455163025e-05,
+      "loss": 3.9286,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.659475498292346e-05,
+      "loss": 3.9209,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.658636903541294e-05,
+      "loss": 3.9322,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657798308790242e-05,
+      "loss": 3.9183,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.65695971403919e-05,
+      "loss": 3.9278,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656121119288138e-05,
+      "loss": 3.9337,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.655284162417459e-05,
+      "loss": 3.919,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.654445567666407e-05,
+      "loss": 3.9415,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.653606972915355e-05,
+      "loss": 3.9436,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.652768378164303e-05,
+      "loss": 3.9357,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.651929783413251e-05,
+      "loss": 3.9268,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.651091188662199e-05,
+      "loss": 3.9235,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6502542317915203e-05,
+      "loss": 3.9289,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6494156370404683e-05,
+      "loss": 3.9224,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6485770422894163e-05,
+      "loss": 3.9374,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6477384475383643e-05,
+      "loss": 3.9196,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6469014906676852e-05,
+      "loss": 3.9324,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460628959166335e-05,
+      "loss": 3.9338,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452243011655815e-05,
+      "loss": 3.9304,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6443857064145295e-05,
+      "loss": 3.9225,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6435471116634775e-05,
+      "loss": 3.9292,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6427085169124255e-05,
+      "loss": 3.9281,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6418699221613735e-05,
+      "loss": 3.9324,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6410329652906944e-05,
+      "loss": 3.9348,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6401943705396424e-05,
+      "loss": 3.9406,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6393557757885908e-05,
+      "loss": 3.9192,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385171810375387e-05,
+      "loss": 3.9354,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.637678586286486e-05,
+      "loss": 3.9357,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368432672961806e-05,
+      "loss": 3.9287,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6360046725451286e-05,
+      "loss": 3.9295,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6351677156744498e-05,
+      "loss": 3.9395,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343291209233978e-05,
+      "loss": 3.9305,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334905261723458e-05,
+      "loss": 3.945,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6326519314212938e-05,
+      "loss": 3.9259,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6318133366702418e-05,
+      "loss": 3.9362,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309747419191898e-05,
+      "loss": 3.9285,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301361471681378e-05,
+      "loss": 3.9245,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.629297552417086e-05,
+      "loss": 3.9267,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284589576660334e-05,
+      "loss": 3.9231,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276203629149814e-05,
+      "loss": 3.9233,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6267817681639294e-05,
+      "loss": 3.9292,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259431734128774e-05,
+      "loss": 3.9242,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.008038520812988,
+      "eval_runtime": 357.5654,
+      "eval_samples_per_second": 1067.192,
+      "eval_steps_per_second": 33.351,
+      "step": 1450064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6251045786618254e-05,
+      "loss": 3.9342,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6242659839107738e-05,
+      "loss": 3.9229,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6234273891597217e-05,
+      "loss": 3.9388,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6225887944086697e-05,
+      "loss": 3.9251,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6217518375379907e-05,
+      "loss": 3.9494,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6209148806673116e-05,
+      "loss": 3.923,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.62007628591626e-05,
+      "loss": 3.9295,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.619237691165208e-05,
+      "loss": 3.9176,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.618399096414156e-05,
+      "loss": 3.9351,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.617560501663104e-05,
+      "loss": 3.9306,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6167235447924248e-05,
+      "loss": 3.9244,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6158849500413728e-05,
+      "loss": 3.9366,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.615047993170694e-05,
+      "loss": 3.9173,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.614209398419642e-05,
+      "loss": 3.9327,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.61337080366859e-05,
+      "loss": 3.9183,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.612532208917538e-05,
+      "loss": 3.9183,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.611693614166486e-05,
+      "loss": 3.9208,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.610855019415434e-05,
+      "loss": 3.9258,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.610016424664382e-05,
+      "loss": 3.9251,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.60917782991333e-05,
+      "loss": 3.943,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6083392351622783e-05,
+      "loss": 3.9245,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6075006404112263e-05,
+      "loss": 3.9332,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6066620456601743e-05,
+      "loss": 3.931,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6058234509091223e-05,
+      "loss": 3.9343,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6049864940384432e-05,
+      "loss": 3.9257,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6041478992873912e-05,
+      "loss": 3.9303,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6033093045363392e-05,
+      "loss": 3.9217,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.602470709785287e-05,
+      "loss": 3.9208,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6016337529146085e-05,
+      "loss": 3.9162,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6007951581635565e-05,
+      "loss": 3.9304,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5999565634125045e-05,
+      "loss": 3.9327,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599117968661452e-05,
+      "loss": 3.9269,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5982810117907737e-05,
+      "loss": 3.9407,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5974424170397217e-05,
+      "loss": 3.9277,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5966054601690426e-05,
+      "loss": 3.9281,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957668654179906e-05,
+      "loss": 3.9314,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5949282706669386e-05,
+      "loss": 3.9287,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5940896759158866e-05,
+      "loss": 3.9165,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5932510811648343e-05,
+      "loss": 3.9115,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5924124864137823e-05,
+      "loss": 3.9209,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5915738916627302e-05,
+      "loss": 3.9253,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5907352969116782e-05,
+      "loss": 3.9326,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5898967021606262e-05,
+      "loss": 3.9253,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5890581074095742e-05,
+      "loss": 3.9178,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5882195126585222e-05,
+      "loss": 3.9342,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5873809179074702e-05,
+      "loss": 3.9221,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5865423231564186e-05,
+      "loss": 3.924,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5857053662857395e-05,
+      "loss": 3.9162,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5848667715346875e-05,
+      "loss": 3.909,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5840281767836355e-05,
+      "loss": 3.9432,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5831895820325835e-05,
+      "loss": 3.932,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5823526251619047e-05,
+      "loss": 3.9237,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5815140304108527e-05,
+      "loss": 3.925,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5806754356598007e-05,
+      "loss": 3.9288,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5798368409087487e-05,
+      "loss": 3.9089,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5789998840380696e-05,
+      "loss": 3.9309,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781612892870176e-05,
+      "loss": 3.9143,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5773226945359656e-05,
+      "loss": 3.9287,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.576484099784914e-05,
+      "loss": 3.9376,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.575645505033862e-05,
+      "loss": 3.9189,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.57480691028281e-05,
+      "loss": 3.9124,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.573968315531758e-05,
+      "loss": 3.9387,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731313586610788e-05,
+      "loss": 3.9025,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5722927639100268e-05,
+      "loss": 3.9165,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5714541691589748e-05,
+      "loss": 3.9317,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.570615574407923e-05,
+      "loss": 3.9306,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.569778617537244e-05,
+      "loss": 3.9218,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.568940022786192e-05,
+      "loss": 3.9126,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.56810142803514e-05,
+      "loss": 3.9105,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.567262833284088e-05,
+      "loss": 3.9124,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5664242385330357e-05,
+      "loss": 3.9313,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5655856437819837e-05,
+      "loss": 3.9142,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5647470490309317e-05,
+      "loss": 3.9285,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5639084542798797e-05,
+      "loss": 3.928,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5630714974092006e-05,
+      "loss": 3.9348,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5622329026581486e-05,
+      "loss": 3.9171,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5613959457874702e-05,
+      "loss": 3.9173,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.560557351036418e-05,
+      "loss": 3.9301,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.559718756285366e-05,
+      "loss": 3.9199,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5588801615343138e-05,
+      "loss": 3.9279,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5580432046636354e-05,
+      "loss": 3.9201,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.557204609912583e-05,
+      "loss": 3.9216,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.556366015161531e-05,
+      "loss": 3.9277,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.555527420410479e-05,
+      "loss": 3.9093,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5546904635398e-05,
+      "loss": 3.9215,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553851868788748e-05,
+      "loss": 3.9198,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553013274037696e-05,
+      "loss": 3.9289,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.552174679286644e-05,
+      "loss": 3.9196,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5513360845355923e-05,
+      "loss": 3.9194,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5504974897845403e-05,
+      "loss": 3.9147,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5496588950334883e-05,
+      "loss": 3.915,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5488203002824363e-05,
+      "loss": 3.9132,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5479817055313843e-05,
+      "loss": 3.9209,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5471447486607052e-05,
+      "loss": 3.9333,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5463061539096532e-05,
+      "loss": 3.9308,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5454675591586015e-05,
+      "loss": 3.9238,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5446289644075495e-05,
+      "loss": 3.9144,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5437903696564975e-05,
+      "loss": 3.9193,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5429534127858184e-05,
+      "loss": 3.921,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5421148180347664e-05,
+      "loss": 3.9291,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5412762232837144e-05,
+      "loss": 3.9276,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5404376285326624e-05,
+      "loss": 3.9209,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5395990337816107e-05,
+      "loss": 3.9149,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5387604390305587e-05,
+      "loss": 3.9351,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5379218442795067e-05,
+      "loss": 3.9089,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.537083249528454e-05,
+      "loss": 3.923,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5362462926577756e-05,
+      "loss": 3.9164,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5354076979067236e-05,
+      "loss": 3.9239,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5345691031556716e-05,
+      "loss": 3.9141,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5337305084046193e-05,
+      "loss": 3.9241,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.532893551533941e-05,
+      "loss": 3.917,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.532054956782889e-05,
+      "loss": 3.9141,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5312179999122098e-05,
+      "loss": 3.9261,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5303794051611578e-05,
+      "loss": 3.9121,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5295408104101058e-05,
+      "loss": 3.9334,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.528702215659054e-05,
+      "loss": 3.9375,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5278636209080014e-05,
+      "loss": 3.9272,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5270250261569494e-05,
+      "loss": 3.9154,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5261864314058974e-05,
+      "loss": 3.9183,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5253478366548454e-05,
+      "loss": 3.9161,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5245108797841667e-05,
+      "loss": 3.9173,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5236722850331146e-05,
+      "loss": 3.9283,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5228336902820626e-05,
+      "loss": 3.9109,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5219950955310106e-05,
+      "loss": 3.9249,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5211581386603315e-05,
+      "loss": 3.9236,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.520321181789653e-05,
+      "loss": 3.9235,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.519482587038601e-05,
+      "loss": 3.9168,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5186439922875488e-05,
+      "loss": 3.9205,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5178053975364968e-05,
+      "loss": 3.9204,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5169668027854448e-05,
+      "loss": 3.9263,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5161282080343928e-05,
+      "loss": 3.9267,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5152896132833408e-05,
+      "loss": 3.9332,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5144510185322888e-05,
+      "loss": 3.9148,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.51361406166161e-05,
+      "loss": 3.9269,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.512775466910558e-05,
+      "loss": 3.9276,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511938510039879e-05,
+      "loss": 3.9181,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511099915288827e-05,
+      "loss": 3.9213,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.510261320537775e-05,
+      "loss": 3.9311,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5094227257867232e-05,
+      "loss": 3.9207,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5085841310356712e-05,
+      "loss": 3.937,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5077455362846192e-05,
+      "loss": 3.9172,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5069069415335672e-05,
+      "loss": 3.926,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5060683467825152e-05,
+      "loss": 3.921,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.505231389911836e-05,
+      "loss": 3.9196,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.504392795160784e-05,
+      "loss": 3.9173,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5035542004097325e-05,
+      "loss": 3.9186,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5027156056586805e-05,
+      "loss": 3.915,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5018786487880014e-05,
+      "loss": 3.9202,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5010400540369494e-05,
+      "loss": 3.9185,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.005103588104248,
+      "eval_runtime": 294.9779,
+      "eval_samples_per_second": 1293.626,
+      "eval_steps_per_second": 40.427,
+      "step": 1526384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5002014592858974e-05,
+      "loss": 3.9234,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4993628645348454e-05,
+      "loss": 3.9158,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4985242697837934e-05,
+      "loss": 3.9287,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4976856750327414e-05,
+      "loss": 3.9197,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4968470802816893e-05,
+      "loss": 3.9387,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4960084855306373e-05,
+      "loss": 3.9219,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4951698907795853e-05,
+      "loss": 3.9211,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4943312960285333e-05,
+      "loss": 3.9082,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4934927012774813e-05,
+      "loss": 3.9228,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4926541065264293e-05,
+      "loss": 3.925,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4918155117753773e-05,
+      "loss": 3.9198,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4909769170243253e-05,
+      "loss": 3.9226,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4901399601536466e-05,
+      "loss": 3.9137,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4893013654025942e-05,
+      "loss": 3.9235,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4884627706515422e-05,
+      "loss": 3.9098,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4876241759004902e-05,
+      "loss": 3.9124,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4867855811494385e-05,
+      "loss": 3.9148,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4859469863983865e-05,
+      "loss": 3.9211,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4851083916473345e-05,
+      "loss": 3.9188,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4842697968962822e-05,
+      "loss": 3.934,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4834312021452302e-05,
+      "loss": 3.9191,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4825926073941782e-05,
+      "loss": 3.9291,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4817540126431262e-05,
+      "loss": 3.9178,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4809170557724474e-05,
+      "loss": 3.9266,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4800784610213954e-05,
+      "loss": 3.9173,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4792398662703434e-05,
+      "loss": 3.9267,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4784012715192914e-05,
+      "loss": 3.9121,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4775626767682394e-05,
+      "loss": 3.9164,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4767240820171874e-05,
+      "loss": 3.9118,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4758854872661354e-05,
+      "loss": 3.9197,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4750468925150834e-05,
+      "loss": 3.9244,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4742099356444046e-05,
+      "loss": 3.9237,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4733729787737256e-05,
+      "loss": 3.9326,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4725343840226736e-05,
+      "loss": 3.919,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4716974271519948e-05,
+      "loss": 3.9192,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4708588324009428e-05,
+      "loss": 3.9256,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4700202376498908e-05,
+      "loss": 3.9234,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4691816428988388e-05,
+      "loss": 3.9081,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4683430481477868e-05,
+      "loss": 3.9076,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4675060912771077e-05,
+      "loss": 3.9099,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4666674965260557e-05,
+      "loss": 3.917,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.465828901775004e-05,
+      "loss": 3.9271,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464990307023952e-05,
+      "loss": 3.919,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464153350153273e-05,
+      "loss": 3.9081,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.463314755402221e-05,
+      "loss": 3.9227,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.462476160651169e-05,
+      "loss": 3.9182,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.461637565900117e-05,
+      "loss": 3.9186,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.460798971149065e-05,
+      "loss": 3.9103,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.459960376398013e-05,
+      "loss": 3.8974,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.459121781646961e-05,
+      "loss": 3.9398,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.458284824776282e-05,
+      "loss": 3.9241,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.457447867905603e-05,
+      "loss": 3.9209,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.456609273154551e-05,
+      "loss": 3.9193,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557706784034994e-05,
+      "loss": 3.9188,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549320836524474e-05,
+      "loss": 3.9076,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.454093488901395e-05,
+      "loss": 3.9194,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.453254894150343e-05,
+      "loss": 3.9108,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.452416299399291e-05,
+      "loss": 3.9172,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.451577704648239e-05,
+      "loss": 3.9326,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.450739109897187e-05,
+      "loss": 3.9121,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.449900515146135e-05,
+      "loss": 3.9038,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4490619203950834e-05,
+      "loss": 3.9294,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.448223325644031e-05,
+      "loss": 3.8985,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.447384730892979e-05,
+      "loss": 3.9068,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.446546136141927e-05,
+      "loss": 3.9255,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.445707541390875e-05,
+      "loss": 3.9195,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.444868946639823e-05,
+      "loss": 3.9155,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.444030351888771e-05,
+      "loss": 3.9073,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4431933950180922e-05,
+      "loss": 3.8983,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4423548002670402e-05,
+      "loss": 3.9067,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4415162055159882e-05,
+      "loss": 3.9235,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4406776107649362e-05,
+      "loss": 3.9093,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.439840653894257e-05,
+      "loss": 3.9181,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.439002059143205e-05,
+      "loss": 3.9218,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4381634643921535e-05,
+      "loss": 3.9297,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373248696411015e-05,
+      "loss": 3.9104,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4364879127704224e-05,
+      "loss": 3.9094,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4356493180193704e-05,
+      "loss": 3.9228,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4348107232683184e-05,
+      "loss": 3.9089,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4339737663976396e-05,
+      "loss": 3.9242,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4331351716465876e-05,
+      "loss": 3.915,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4322965768955356e-05,
+      "loss": 3.9137,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4314596200248565e-05,
+      "loss": 3.921,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4306210252738045e-05,
+      "loss": 3.9041,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4297824305227525e-05,
+      "loss": 3.9129,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4289438357717005e-05,
+      "loss": 3.9117,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4281052410206488e-05,
+      "loss": 3.9224,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4272666462695965e-05,
+      "loss": 3.9109,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4264280515185445e-05,
+      "loss": 3.917,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4255894567674925e-05,
+      "loss": 3.9062,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4247524998968134e-05,
+      "loss": 3.9071,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4239139051457617e-05,
+      "loss": 3.905,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4230753103947097e-05,
+      "loss": 3.9131,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4222367156436577e-05,
+      "loss": 3.9266,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4213981208926057e-05,
+      "loss": 3.9214,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205595261415537e-05,
+      "loss": 3.9206,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4197209313905017e-05,
+      "loss": 3.9081,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4188823366394494e-05,
+      "loss": 3.9105,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4180453797687706e-05,
+      "loss": 3.9138,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.417206785017719e-05,
+      "loss": 3.922,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.416368190266667e-05,
+      "loss": 3.9201,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4155295955156146e-05,
+      "loss": 3.9164,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.414692638644936e-05,
+      "loss": 3.9059,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.413854043893884e-05,
+      "loss": 3.9288,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4130154491428318e-05,
+      "loss": 3.903,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4121768543917798e-05,
+      "loss": 3.9172,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4113382596407278e-05,
+      "loss": 3.9089,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4104996648896758e-05,
+      "loss": 3.9134,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4096610701386238e-05,
+      "loss": 3.9068,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4088241132679447e-05,
+      "loss": 3.9171,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4079855185168927e-05,
+      "loss": 3.908,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.407146923765841e-05,
+      "loss": 3.9101,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.406308329014789e-05,
+      "loss": 3.9145,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.405469734263737e-05,
+      "loss": 3.9096,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4046311395126847e-05,
+      "loss": 3.9217,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4037925447616327e-05,
+      "loss": 3.9299,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.402955587890954e-05,
+      "loss": 3.9152,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.402116993139902e-05,
+      "loss": 3.9118,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.40127839838885e-05,
+      "loss": 3.9121,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.400439803637798e-05,
+      "loss": 3.9121,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3996028467671192e-05,
+      "loss": 3.909,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3987642520160672e-05,
+      "loss": 3.9218,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.397925657265015e-05,
+      "loss": 3.8997,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3970887003943364e-05,
+      "loss": 3.9178,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3962501056432844e-05,
+      "loss": 3.9116,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.395411510892232e-05,
+      "loss": 3.9201,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.39457291614118e-05,
+      "loss": 3.9084,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.393734321390128e-05,
+      "loss": 3.9156,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392895726639076e-05,
+      "loss": 3.9084,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392057131888024e-05,
+      "loss": 3.9231,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3912201750173453e-05,
+      "loss": 3.9187,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3903815802662933e-05,
+      "loss": 3.9254,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3895429855152413e-05,
+      "loss": 3.9055,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3887043907641893e-05,
+      "loss": 3.9211,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3878657960131373e-05,
+      "loss": 3.921,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3870288391424582e-05,
+      "loss": 3.9116,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3861902443914065e-05,
+      "loss": 3.9138,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3853516496403545e-05,
+      "loss": 3.9239,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3845130548893025e-05,
+      "loss": 3.9165,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3836744601382502e-05,
+      "loss": 3.9313,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3828358653871982e-05,
+      "loss": 3.9094,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381997270636146e-05,
+      "loss": 3.9214,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381158675885094e-05,
+      "loss": 3.9122,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3803217190144154e-05,
+      "loss": 3.9123,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3794847621437367e-05,
+      "loss": 3.916,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3786461673926847e-05,
+      "loss": 3.9058,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3778075726416327e-05,
+      "loss": 3.9081,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3769689778905803e-05,
+      "loss": 3.9088,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3761303831395283e-05,
+      "loss": 3.9133,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.001642227172852,
+      "eval_runtime": 374.1681,
+      "eval_samples_per_second": 1019.838,
+      "eval_steps_per_second": 31.871,
+      "step": 1602704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3752917883884766e-05,
+      "loss": 3.9151,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3744531936374246e-05,
+      "loss": 3.911,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3736145988863726e-05,
+      "loss": 3.9192,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3727760041353206e-05,
+      "loss": 3.9172,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3719374093842683e-05,
+      "loss": 3.9292,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3710988146332163e-05,
+      "loss": 3.9153,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3702602198821643e-05,
+      "loss": 3.9088,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3694216251311123e-05,
+      "loss": 3.9052,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3685830303800606e-05,
+      "loss": 3.914,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3677444356290086e-05,
+      "loss": 3.9149,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3669058408779566e-05,
+      "loss": 3.916,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3660672461269042e-05,
+      "loss": 3.9135,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3652302892562255e-05,
+      "loss": 3.9087,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3643916945051735e-05,
+      "loss": 3.9134,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3635530997541215e-05,
+      "loss": 3.9056,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3627145050030695e-05,
+      "loss": 3.9007,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3618759102520175e-05,
+      "loss": 3.9089,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3610373155009655e-05,
+      "loss": 3.9136,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3601987207499135e-05,
+      "loss": 3.9114,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3593601259988615e-05,
+      "loss": 3.9277,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3585215312478095e-05,
+      "loss": 3.9144,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3576829364967575e-05,
+      "loss": 3.9232,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3568443417457055e-05,
+      "loss": 3.9079,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3560057469946534e-05,
+      "loss": 3.9177,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3551687901239747e-05,
+      "loss": 3.9123,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3543301953729224e-05,
+      "loss": 3.9193,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3534932385022436e-05,
+      "loss": 3.9089,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3526546437511916e-05,
+      "loss": 3.9045,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.35181604900014e-05,
+      "loss": 3.9057,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3509774542490876e-05,
+      "loss": 3.9127,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3501388594980356e-05,
+      "loss": 3.914,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3493002647469836e-05,
+      "loss": 3.9162,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3484616699959316e-05,
+      "loss": 3.9224,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3476247131252528e-05,
+      "loss": 3.9137,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3467861183742008e-05,
+      "loss": 3.9113,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.345949161503522e-05,
+      "loss": 3.9233,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3451105667524697e-05,
+      "loss": 3.9153,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3442719720014177e-05,
+      "loss": 3.8999,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3434333772503657e-05,
+      "loss": 3.9024,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3425947824993137e-05,
+      "loss": 3.9023,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.341756187748262e-05,
+      "loss": 3.9074,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.34091759299721e-05,
+      "loss": 3.9225,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3400789982461577e-05,
+      "loss": 3.9112,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.339243679255852e-05,
+      "loss": 3.9041,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3384050845048e-05,
+      "loss": 3.9169,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3375664897537482e-05,
+      "loss": 3.9044,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3367278950026962e-05,
+      "loss": 3.9128,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3358893002516442e-05,
+      "loss": 3.9022,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3350507055005922e-05,
+      "loss": 3.8854,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3342121107495402e-05,
+      "loss": 3.9356,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3333735159984878e-05,
+      "loss": 3.9125,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3325349212474358e-05,
+      "loss": 3.9167,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3316979643767574e-05,
+      "loss": 3.9097,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.330859369625705e-05,
+      "loss": 3.9137,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.330020774874653e-05,
+      "loss": 3.9009,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.329182180123601e-05,
+      "loss": 3.9057,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.328343585372549e-05,
+      "loss": 3.9062,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.327504990621497e-05,
+      "loss": 3.9118,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3266680337508183e-05,
+      "loss": 3.9245,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3258294389997663e-05,
+      "loss": 3.9046,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3249908442487143e-05,
+      "loss": 3.8953,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3241522494976623e-05,
+      "loss": 3.9229,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3233136547466103e-05,
+      "loss": 3.8911,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3224750599955583e-05,
+      "loss": 3.8976,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3216381031248792e-05,
+      "loss": 3.9149,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3207995083738275e-05,
+      "loss": 3.9194,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3199609136227755e-05,
+      "loss": 3.9063,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3191223188717232e-05,
+      "loss": 3.9056,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3182837241206712e-05,
+      "loss": 3.8868,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.317445129369619e-05,
+      "loss": 3.9025,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3166081724989404e-05,
+      "loss": 3.9138,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3157695777478884e-05,
+      "loss": 3.9008,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3149309829968364e-05,
+      "loss": 3.9102,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3140923882457844e-05,
+      "loss": 3.9128,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3132537934947324e-05,
+      "loss": 3.9233,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3124151987436804e-05,
+      "loss": 3.9018,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3115766039926284e-05,
+      "loss": 3.9062,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3107380092415764e-05,
+      "loss": 3.9135,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3099026902512705e-05,
+      "loss": 3.8999,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3090640955002185e-05,
+      "loss": 3.915,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3082255007491665e-05,
+      "loss": 3.9075,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3073869059981145e-05,
+      "loss": 3.9089,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3065483112470625e-05,
+      "loss": 3.9164,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3057113543763838e-05,
+      "loss": 3.8975,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3048727596253318e-05,
+      "loss": 3.8991,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3040341648742798e-05,
+      "loss": 3.9073,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3031955701232278e-05,
+      "loss": 3.9163,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3023569753721758e-05,
+      "loss": 3.8989,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3015183806211238e-05,
+      "loss": 3.912,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3006814237504447e-05,
+      "loss": 3.8985,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.299842828999393e-05,
+      "loss": 3.9006,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.299004234248341e-05,
+      "loss": 3.8987,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2981656394972887e-05,
+      "loss": 3.905,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2973270447462366e-05,
+      "loss": 3.9192,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2964884499951846e-05,
+      "loss": 3.9132,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2956498552441326e-05,
+      "loss": 3.9139,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2948112604930806e-05,
+      "loss": 3.9018,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.293975941502775e-05,
+      "loss": 3.9023,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.293137346751723e-05,
+      "loss": 3.9071,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.292298752000671e-05,
+      "loss": 3.9144,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2914601572496188e-05,
+      "loss": 3.9092,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2906215624985668e-05,
+      "loss": 3.9084,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2897829677475148e-05,
+      "loss": 3.906,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.288944372996463e-05,
+      "loss": 3.919,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.288105778245411e-05,
+      "loss": 3.8942,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.287268821374732e-05,
+      "loss": 3.9096,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.28643022662368e-05,
+      "loss": 3.9004,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.285591631872628e-05,
+      "loss": 3.9111,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2847546750019493e-05,
+      "loss": 3.8998,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2839160802508972e-05,
+      "loss": 3.9096,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2830774854998452e-05,
+      "loss": 3.9003,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2822388907487932e-05,
+      "loss": 3.9038,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2814002959977412e-05,
+      "loss": 3.9067,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.280563339127062e-05,
+      "loss": 3.9018,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2797263822563834e-05,
+      "loss": 3.9154,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2788877875053314e-05,
+      "loss": 3.9235,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2780491927542794e-05,
+      "loss": 3.9104,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2772105980032274e-05,
+      "loss": 3.9105,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2763720032521754e-05,
+      "loss": 3.9014,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2755334085011234e-05,
+      "loss": 3.9038,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2746948137500714e-05,
+      "loss": 3.908,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2738562189990194e-05,
+      "loss": 3.9175,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2730176242479674e-05,
+      "loss": 3.8912,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2721790294969154e-05,
+      "loss": 3.9099,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2713404347458633e-05,
+      "loss": 3.9061,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2705018399948113e-05,
+      "loss": 3.9111,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2696648831241323e-05,
+      "loss": 3.9037,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268827926253454e-05,
+      "loss": 3.9055,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2679893315024015e-05,
+      "loss": 3.9049,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2671507367513495e-05,
+      "loss": 3.9133,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2663121420002975e-05,
+      "loss": 3.916,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2654751851296187e-05,
+      "loss": 3.9163,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2646365903785667e-05,
+      "loss": 3.9001,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2637979956275147e-05,
+      "loss": 3.9142,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2629594008764627e-05,
+      "loss": 3.9132,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2621208061254107e-05,
+      "loss": 3.9058,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2612822113743587e-05,
+      "loss": 3.9087,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2604436166233067e-05,
+      "loss": 3.9155,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2596066597526276e-05,
+      "loss": 3.9069,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2587680650015756e-05,
+      "loss": 3.9274,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257929470250524e-05,
+      "loss": 3.9055,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257090875499472e-05,
+      "loss": 3.9093,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2562522807484196e-05,
+      "loss": 3.904,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2554136859973676e-05,
+      "loss": 3.9116,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2545750912463156e-05,
+      "loss": 3.9067,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2537364964952636e-05,
+      "loss": 3.9016,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.252899539624585e-05,
+      "loss": 3.9041,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.252062582753906e-05,
+      "loss": 3.8965,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.251223988002854e-05,
+      "loss": 3.9104,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9993362426757812,
+      "eval_runtime": 294.0271,
+      "eval_samples_per_second": 1297.809,
+      "eval_steps_per_second": 40.557,
+      "step": 1679024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2503853932518017e-05,
+      "loss": 3.9125,
+      "step": 1679360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2495467985007497e-05,
+      "loss": 3.9036,
+      "step": 1679872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2487082037496977e-05,
+      "loss": 3.9093,
+      "step": 1680384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.247869608998646e-05,
+      "loss": 3.9104,
+      "step": 1680896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.247031014247594e-05,
+      "loss": 3.923,
+      "step": 1681408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.246192419496542e-05,
+      "loss": 3.9064,
+      "step": 1681920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.245355462625863e-05,
+      "loss": 3.9068,
+      "step": 1682432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.244516867874811e-05,
+      "loss": 3.8931,
+      "step": 1682944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.243678273123759e-05,
+      "loss": 3.907,
+      "step": 1683456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.242839678372707e-05,
+      "loss": 3.9109,
+      "step": 1683968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2420010836216553e-05,
+      "loss": 3.9092,
+      "step": 1684480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2411641267509762e-05,
+      "loss": 3.9067,
+      "step": 1684992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.240327169880297e-05,
+      "loss": 3.9001,
+      "step": 1685504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.239488575129245e-05,
+      "loss": 3.9074,
+      "step": 1686016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.238649980378193e-05,
+      "loss": 3.9019,
+      "step": 1686528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2378113856271414e-05,
+      "loss": 3.8955,
+      "step": 1687040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2369727908760894e-05,
+      "loss": 3.9013,
+      "step": 1687552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2361341961250374e-05,
+      "loss": 3.9045,
+      "step": 1688064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2352972392543583e-05,
+      "loss": 3.9061,
+      "step": 1688576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2344586445033063e-05,
+      "loss": 3.9157,
+      "step": 1689088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2336200497522543e-05,
+      "loss": 3.9118,
+      "step": 1689600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2327814550012023e-05,
+      "loss": 3.9158,
+      "step": 1690112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2319428602501503e-05,
+      "loss": 3.9008,
+      "step": 1690624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2311042654990983e-05,
+      "loss": 3.9106,
+      "step": 1691136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2302656707480463e-05,
+      "loss": 3.9073,
+      "step": 1691648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2294270759969943e-05,
+      "loss": 3.9128,
+      "step": 1692160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2285901191263152e-05,
+      "loss": 3.8986,
+      "step": 1692672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2277515243752632e-05,
+      "loss": 3.9023,
+      "step": 1693184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2269129296242115e-05,
+      "loss": 3.9014,
+      "step": 1693696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2260759727535324e-05,
+      "loss": 3.9053,
+      "step": 1694208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2252373780024804e-05,
+      "loss": 3.9061,
+      "step": 1694720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2243987832514284e-05,
+      "loss": 3.915,
+      "step": 1695232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2235601885003764e-05,
+      "loss": 3.9124,
+      "step": 1695744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2227215937493244e-05,
+      "loss": 3.9079,
+      "step": 1696256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2218846368786457e-05,
+      "loss": 3.9082,
+      "step": 1696768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.221047680007967e-05,
+      "loss": 3.9109,
+      "step": 1697280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2202090852569146e-05,
+      "loss": 3.9084,
+      "step": 1697792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2193704905058626e-05,
+      "loss": 3.8982,
+      "step": 1698304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2185318957548106e-05,
+      "loss": 3.8953,
+      "step": 1698816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2176933010037586e-05,
+      "loss": 3.8961,
+      "step": 1699328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.216854706252707e-05,
+      "loss": 3.8991,
+      "step": 1699840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.216016111501655e-05,
+      "loss": 3.9129,
+      "step": 1700352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.215177516750603e-05,
+      "loss": 3.907,
+      "step": 1700864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2143405598799238e-05,
+      "loss": 3.9005,
+      "step": 1701376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2135019651288718e-05,
+      "loss": 3.905,
+      "step": 1701888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2126633703778198e-05,
+      "loss": 3.9013,
+      "step": 1702400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2118247756267678e-05,
+      "loss": 3.9071,
+      "step": 1702912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2109861808757158e-05,
+      "loss": 3.8955,
+      "step": 1703424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2101475861246638e-05,
+      "loss": 3.8795,
+      "step": 1703936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2093089913736118e-05,
+      "loss": 3.9263,
+      "step": 1704448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2084703966225598e-05,
+      "loss": 3.9041,
+      "step": 1704960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2076318018715078e-05,
+      "loss": 3.9142,
+      "step": 1705472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2067948450008287e-05,
+      "loss": 3.9017,
+      "step": 1705984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.205956250249777e-05,
+      "loss": 3.9042,
+      "step": 1706496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.205117655498725e-05,
+      "loss": 3.8963,
+      "step": 1707008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.204280698628046e-05,
+      "loss": 3.9008,
+      "step": 1707520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.203442103876994e-05,
+      "loss": 3.9028,
+      "step": 1708032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.202603509125942e-05,
+      "loss": 3.8989,
+      "step": 1708544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.20176491437489e-05,
+      "loss": 3.9211,
+      "step": 1709056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.200926319623838e-05,
+      "loss": 3.8964,
+      "step": 1709568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.200087724872786e-05,
+      "loss": 3.8913,
+      "step": 1710080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.199249130121734e-05,
+      "loss": 3.9167,
+      "step": 1710592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.198410535370682e-05,
+      "loss": 3.8834,
+      "step": 1711104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.19757194061963e-05,
+      "loss": 3.8898,
+      "step": 1711616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.196736621629324e-05,
+      "loss": 3.9114,
+      "step": 1712128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1958980268782724e-05,
+      "loss": 3.9122,
+      "step": 1712640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1950594321272204e-05,
+      "loss": 3.8956,
+      "step": 1713152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1942208373761684e-05,
+      "loss": 3.9027,
+      "step": 1713664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.193382242625116e-05,
+      "loss": 3.8812,
+      "step": 1714176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.192543647874064e-05,
+      "loss": 3.8988,
+      "step": 1714688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.191705053123012e-05,
+      "loss": 3.9051,
+      "step": 1715200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.19086645837196e-05,
+      "loss": 3.8945,
+      "step": 1715712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1900295015012813e-05,
+      "loss": 3.9034,
+      "step": 1716224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1891909067502293e-05,
+      "loss": 3.9072,
+      "step": 1716736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1883523119991773e-05,
+      "loss": 3.9186,
+      "step": 1717248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1875137172481253e-05,
+      "loss": 3.8963,
+      "step": 1717760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1866751224970732e-05,
+      "loss": 3.8957,
+      "step": 1718272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1858365277460212e-05,
+      "loss": 3.9102,
+      "step": 1718784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1849979329949692e-05,
+      "loss": 3.8966,
+      "step": 1719296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1841593382439172e-05,
+      "loss": 3.9041,
+      "step": 1719808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1833223813732385e-05,
+      "loss": 3.9048,
+      "step": 1720320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1824837866221865e-05,
+      "loss": 3.8987,
+      "step": 1720832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1816468297515074e-05,
+      "loss": 3.9111,
+      "step": 1721344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1808082350004554e-05,
+      "loss": 3.8899,
+      "step": 1721856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1799696402494034e-05,
+      "loss": 3.893,
+      "step": 1722368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1791310454983514e-05,
+      "loss": 3.9029,
+      "step": 1722880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1782924507472994e-05,
+      "loss": 3.909,
+      "step": 1723392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1774538559962474e-05,
+      "loss": 3.8955,
+      "step": 1723904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1766152612451954e-05,
+      "loss": 3.902,
+      "step": 1724416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1757766664941434e-05,
+      "loss": 3.8882,
+      "step": 1724928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1749397096234646e-05,
+      "loss": 3.8982,
+      "step": 1725440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1741011148724126e-05,
+      "loss": 3.8903,
+      "step": 1725952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1732625201213606e-05,
+      "loss": 3.9016,
+      "step": 1726464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1724239253703086e-05,
+      "loss": 3.9148,
+      "step": 1726976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1715869684996295e-05,
+      "loss": 3.9041,
+      "step": 1727488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1707483737485775e-05,
+      "loss": 3.9043,
+      "step": 1728000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1699114168778987e-05,
+      "loss": 3.8942,
+      "step": 1728512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1690728221268467e-05,
+      "loss": 3.8975,
+      "step": 1729024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1682342273757947e-05,
+      "loss": 3.8997,
+      "step": 1729536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1673956326247427e-05,
+      "loss": 3.9093,
+      "step": 1730048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1665570378736907e-05,
+      "loss": 3.903,
+      "step": 1730560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1657184431226387e-05,
+      "loss": 3.9063,
+      "step": 1731072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1648814862519596e-05,
+      "loss": 3.8959,
+      "step": 1731584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.164042891500908e-05,
+      "loss": 3.9105,
+      "step": 1732096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.163204296749856e-05,
+      "loss": 3.8935,
+      "step": 1732608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.162365701998804e-05,
+      "loss": 3.9032,
+      "step": 1733120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1615271072477516e-05,
+      "loss": 3.8902,
+      "step": 1733632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.160690150377073e-05,
+      "loss": 3.907,
+      "step": 1734144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.159851555626021e-05,
+      "loss": 3.8933,
+      "step": 1734656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.159012960874969e-05,
+      "loss": 3.8989,
+      "step": 1735168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.158174366123917e-05,
+      "loss": 3.8961,
+      "step": 1735680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.157335771372865e-05,
+      "loss": 3.8995,
+      "step": 1736192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.156497176621813e-05,
+      "loss": 3.8975,
+      "step": 1736704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.155658581870761e-05,
+      "loss": 3.8941,
+      "step": 1737216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.154819987119709e-05,
+      "loss": 3.9095,
+      "step": 1737728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1539813923686568e-05,
+      "loss": 3.9148,
+      "step": 1738240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.153144435497978e-05,
+      "loss": 3.9058,
+      "step": 1738752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.152307478627299e-05,
+      "loss": 3.9016,
+      "step": 1739264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1514705217566202e-05,
+      "loss": 3.8956,
+      "step": 1739776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1506319270055682e-05,
+      "loss": 3.8972,
+      "step": 1740288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1497933322545162e-05,
+      "loss": 3.9002,
+      "step": 1740800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1489547375034642e-05,
+      "loss": 3.907,
+      "step": 1741312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1481161427524122e-05,
+      "loss": 3.8845,
+      "step": 1741824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1472775480013602e-05,
+      "loss": 3.9044,
+      "step": 1742336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1464389532503082e-05,
+      "loss": 3.9011,
+      "step": 1742848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1456003584992562e-05,
+      "loss": 3.9024,
+      "step": 1743360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1447617637482042e-05,
+      "loss": 3.8953,
+      "step": 1743872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1439248068775254e-05,
+      "loss": 3.9016,
+      "step": 1744384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1430862121264734e-05,
+      "loss": 3.8982,
+      "step": 1744896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1422476173754214e-05,
+      "loss": 3.9082,
+      "step": 1745408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1414090226243694e-05,
+      "loss": 3.9072,
+      "step": 1745920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1405720657536903e-05,
+      "loss": 3.9073,
+      "step": 1746432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1397334710026383e-05,
+      "loss": 3.8946,
+      "step": 1746944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1388948762515863e-05,
+      "loss": 3.9077,
+      "step": 1747456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1380562815005343e-05,
+      "loss": 3.9072,
+      "step": 1747968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1372193246298556e-05,
+      "loss": 3.896,
+      "step": 1748480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1363807298788036e-05,
+      "loss": 3.9042,
+      "step": 1748992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1355421351277516e-05,
+      "loss": 3.9076,
+      "step": 1749504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1347035403766996e-05,
+      "loss": 3.904,
+      "step": 1750016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1338665835060205e-05,
+      "loss": 3.9189,
+      "step": 1750528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1330296266353417e-05,
+      "loss": 3.8977,
+      "step": 1751040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1321910318842897e-05,
+      "loss": 3.9063,
+      "step": 1751552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1313524371332377e-05,
+      "loss": 3.9017,
+      "step": 1752064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1305138423821857e-05,
+      "loss": 3.9016,
+      "step": 1752576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1296752476311337e-05,
+      "loss": 3.8974,
+      "step": 1753088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1288366528800817e-05,
+      "loss": 3.893,
+      "step": 1753600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1279980581290297e-05,
+      "loss": 3.8981,
+      "step": 1754112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.127161101258351e-05,
+      "loss": 3.8927,
+      "step": 1754624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.126322506507299e-05,
+      "loss": 3.9076,
+      "step": 1755136
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.996788501739502,
+      "eval_runtime": 293.3911,
+      "eval_samples_per_second": 1300.622,
+      "eval_steps_per_second": 40.645,
+      "step": 1755344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.125483911756247e-05,
+      "loss": 3.9081,
+      "step": 1755648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1246453170051946e-05,
+      "loss": 3.8962,
+      "step": 1756160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1238067222541426e-05,
+      "loss": 3.9048,
+      "step": 1756672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.122968127503091e-05,
+      "loss": 3.9025,
+      "step": 1757184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.122129532752039e-05,
+      "loss": 3.9117,
+      "step": 1757696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.121290938000987e-05,
+      "loss": 3.9027,
+      "step": 1758208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.120452343249935e-05,
+      "loss": 3.9035,
+      "step": 1758720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1196137484988826e-05,
+      "loss": 3.8926,
+      "step": 1759232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1187751537478306e-05,
+      "loss": 3.8975,
+      "step": 1759744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1179365589967786e-05,
+      "loss": 3.9053,
+      "step": 1760256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1170979642457266e-05,
+      "loss": 3.9003,
+      "step": 1760768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.116259369494675e-05,
+      "loss": 3.8986,
+      "step": 1761280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1154224126239958e-05,
+      "loss": 3.901,
+      "step": 1761792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1145838178729438e-05,
+      "loss": 3.8947,
+      "step": 1762304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1137452231218918e-05,
+      "loss": 3.8933,
+      "step": 1762816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1129066283708398e-05,
+      "loss": 3.8902,
+      "step": 1763328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1120680336197878e-05,
+      "loss": 3.8915,
+      "step": 1763840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1112294388687358e-05,
+      "loss": 3.9024,
+      "step": 1764352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.110392481998057e-05,
+      "loss": 3.8943,
+      "step": 1764864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.109553887247005e-05,
+      "loss": 3.9153,
+      "step": 1765376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.108715292495953e-05,
+      "loss": 3.9034,
+      "step": 1765888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1078766977449007e-05,
+      "loss": 3.9087,
+      "step": 1766400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1070381029938487e-05,
+      "loss": 3.8959,
+      "step": 1766912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1061995082427967e-05,
+      "loss": 3.9016,
+      "step": 1767424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.105360913491745e-05,
+      "loss": 3.9003,
+      "step": 1767936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.104522318740693e-05,
+      "loss": 3.9054,
+      "step": 1768448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.103685361870014e-05,
+      "loss": 3.8957,
+      "step": 1768960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.102846767118962e-05,
+      "loss": 3.8981,
+      "step": 1769472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.10200817236791e-05,
+      "loss": 3.8915,
+      "step": 1769984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.101169577616858e-05,
+      "loss": 3.8981,
+      "step": 1770496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.100330982865806e-05,
+      "loss": 3.8968,
+      "step": 1771008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.099494025995127e-05,
+      "loss": 3.91,
+      "step": 1771520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.098655431244075e-05,
+      "loss": 3.9044,
+      "step": 1772032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.097816836493023e-05,
+      "loss": 3.9051,
+      "step": 1772544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.096978241741971e-05,
+      "loss": 3.9014,
+      "step": 1773056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.096141284871292e-05,
+      "loss": 3.9013,
+      "step": 1773568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0953026901202404e-05,
+      "loss": 3.9053,
+      "step": 1774080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0944640953691884e-05,
+      "loss": 3.8904,
+      "step": 1774592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0936255006181364e-05,
+      "loss": 3.8898,
+      "step": 1775104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.092786905867084e-05,
+      "loss": 3.8841,
+      "step": 1775616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0919499489964053e-05,
+      "loss": 3.8959,
+      "step": 1776128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0911113542453533e-05,
+      "loss": 3.9003,
+      "step": 1776640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0902727594943013e-05,
+      "loss": 3.9006,
+      "step": 1777152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0894341647432492e-05,
+      "loss": 3.8955,
+      "step": 1777664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0885955699921972e-05,
+      "loss": 3.8962,
+      "step": 1778176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0877569752411452e-05,
+      "loss": 3.8972,
+      "step": 1778688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.086920018370466e-05,
+      "loss": 3.9028,
+      "step": 1779200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.086081423619414e-05,
+      "loss": 3.8886,
+      "step": 1779712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.085242828868362e-05,
+      "loss": 3.8742,
+      "step": 1780224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0844042341173105e-05,
+      "loss": 3.9126,
+      "step": 1780736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0835656393662585e-05,
+      "loss": 3.9024,
+      "step": 1781248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0827270446152065e-05,
+      "loss": 3.9069,
+      "step": 1781760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0818900877445274e-05,
+      "loss": 3.8959,
+      "step": 1782272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0810514929934754e-05,
+      "loss": 3.8985,
+      "step": 1782784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0802128982424234e-05,
+      "loss": 3.8889,
+      "step": 1783296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0793743034913714e-05,
+      "loss": 3.8938,
+      "step": 1783808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0785357087403194e-05,
+      "loss": 3.8983,
+      "step": 1784320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0776971139892674e-05,
+      "loss": 3.8904,
+      "step": 1784832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0768585192382153e-05,
+      "loss": 3.9154,
+      "step": 1785344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0760199244871633e-05,
+      "loss": 3.8966,
+      "step": 1785856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0751813297361113e-05,
+      "loss": 3.8803,
+      "step": 1786368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0743427349850593e-05,
+      "loss": 3.9107,
+      "step": 1786880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0735041402340073e-05,
+      "loss": 3.8783,
+      "step": 1787392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0726655454829553e-05,
+      "loss": 3.8806,
+      "step": 1787904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0718285886122766e-05,
+      "loss": 3.9086,
+      "step": 1788416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0709916317415975e-05,
+      "loss": 3.9032,
+      "step": 1788928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0701530369905455e-05,
+      "loss": 3.8878,
+      "step": 1789440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0693144422394935e-05,
+      "loss": 3.8974,
+      "step": 1789952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0684758474884418e-05,
+      "loss": 3.877,
+      "step": 1790464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0676372527373898e-05,
+      "loss": 3.8927,
+      "step": 1790976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0668002958667107e-05,
+      "loss": 3.8945,
+      "step": 1791488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0659617011156587e-05,
+      "loss": 3.8905,
+      "step": 1792000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0651231063646067e-05,
+      "loss": 3.9016,
+      "step": 1792512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0642845116135547e-05,
+      "loss": 3.8961,
+      "step": 1793024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.063447554742876e-05,
+      "loss": 3.9081,
+      "step": 1793536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.062608959991824e-05,
+      "loss": 3.8914,
+      "step": 1794048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.061770365240772e-05,
+      "loss": 3.8937,
+      "step": 1794560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.06093177048972e-05,
+      "loss": 3.9004,
+      "step": 1795072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0600931757386676e-05,
+      "loss": 3.8886,
+      "step": 1795584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0592545809876156e-05,
+      "loss": 3.9005,
+      "step": 1796096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.058417624116937e-05,
+      "loss": 3.8951,
+      "step": 1796608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.057579029365885e-05,
+      "loss": 3.8936,
+      "step": 1797120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0567404346148328e-05,
+      "loss": 3.9085,
+      "step": 1797632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0559018398637808e-05,
+      "loss": 3.8798,
+      "step": 1798144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.055064882993102e-05,
+      "loss": 3.8849,
+      "step": 1798656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0542262882420497e-05,
+      "loss": 3.9005,
+      "step": 1799168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.053387693490998e-05,
+      "loss": 3.9,
+      "step": 1799680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.052549098739946e-05,
+      "loss": 3.8909,
+      "step": 1800192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.051710503988894e-05,
+      "loss": 3.8997,
+      "step": 1800704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.050871909237842e-05,
+      "loss": 3.8801,
+      "step": 1801216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.05003331448679e-05,
+      "loss": 3.8904,
+      "step": 1801728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0491947197357377e-05,
+      "loss": 3.8887,
+      "step": 1802240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0483561249846857e-05,
+      "loss": 3.8938,
+      "step": 1802752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0475191681140073e-05,
+      "loss": 3.908,
+      "step": 1803264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0466805733629553e-05,
+      "loss": 3.8975,
+      "step": 1803776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.045841978611903e-05,
+      "loss": 3.8996,
+      "step": 1804288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.045003383860851e-05,
+      "loss": 3.8889,
+      "step": 1804800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.044164789109799e-05,
+      "loss": 3.8898,
+      "step": 1805312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0433278322391202e-05,
+      "loss": 3.8928,
+      "step": 1805824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0424892374880682e-05,
+      "loss": 3.9021,
+      "step": 1806336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.041650642737016e-05,
+      "loss": 3.8967,
+      "step": 1806848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.040812047985964e-05,
+      "loss": 3.9028,
+      "step": 1807360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.039973453234912e-05,
+      "loss": 3.8888,
+      "step": 1807872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.03913485848386e-05,
+      "loss": 3.9014,
+      "step": 1808384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.038296263732808e-05,
+      "loss": 3.891,
+      "step": 1808896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0374576689817558e-05,
+      "loss": 3.8965,
+      "step": 1809408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0366207121110774e-05,
+      "loss": 3.8823,
+      "step": 1809920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0357821173600254e-05,
+      "loss": 3.9005,
+      "step": 1810432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0349435226089734e-05,
+      "loss": 3.8868,
+      "step": 1810944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.034104927857921e-05,
+      "loss": 3.8961,
+      "step": 1811456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.033266333106869e-05,
+      "loss": 3.8864,
+      "step": 1811968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.032427738355817e-05,
+      "loss": 3.8975,
+      "step": 1812480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0315907814851383e-05,
+      "loss": 3.8913,
+      "step": 1812992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0307521867340863e-05,
+      "loss": 3.8865,
+      "step": 1813504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0299135919830343e-05,
+      "loss": 3.9048,
+      "step": 1814016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0290749972319823e-05,
+      "loss": 3.909,
+      "step": 1814528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0282380403613032e-05,
+      "loss": 3.902,
+      "step": 1815040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0273994456102512e-05,
+      "loss": 3.8953,
+      "step": 1815552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.026560850859199e-05,
+      "loss": 3.887,
+      "step": 1816064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0257222561081475e-05,
+      "loss": 3.8911,
+      "step": 1816576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0248852992374684e-05,
+      "loss": 3.8964,
+      "step": 1817088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0240467044864164e-05,
+      "loss": 3.8974,
+      "step": 1817600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0232097476157377e-05,
+      "loss": 3.8823,
+      "step": 1818112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0223711528646857e-05,
+      "loss": 3.8991,
+      "step": 1818624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0215325581136336e-05,
+      "loss": 3.8983,
+      "step": 1819136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0206939633625816e-05,
+      "loss": 3.8953,
+      "step": 1819648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0198553686115296e-05,
+      "loss": 3.8898,
+      "step": 1820160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0190184117408505e-05,
+      "loss": 3.8931,
+      "step": 1820672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0181798169897985e-05,
+      "loss": 3.8894,
+      "step": 1821184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0173412222387465e-05,
+      "loss": 3.9043,
+      "step": 1821696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0165026274876945e-05,
+      "loss": 3.9027,
+      "step": 1822208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.015664032736643e-05,
+      "loss": 3.8997,
+      "step": 1822720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.014825437985591e-05,
+      "loss": 3.8848,
+      "step": 1823232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.013986843234539e-05,
+      "loss": 3.9049,
+      "step": 1823744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0131482484834865e-05,
+      "loss": 3.9013,
+      "step": 1824256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0123096537324345e-05,
+      "loss": 3.8897,
+      "step": 1824768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.011474334742129e-05,
+      "loss": 3.9004,
+      "step": 1825280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.010635739991077e-05,
+      "loss": 3.8959,
+      "step": 1825792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.009797145240025e-05,
+      "loss": 3.9004,
+      "step": 1826304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.008960188369346e-05,
+      "loss": 3.9095,
+      "step": 1826816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.008121593618294e-05,
+      "loss": 3.8918,
+      "step": 1827328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.007282998867242e-05,
+      "loss": 3.8984,
+      "step": 1827840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.00644440411619e-05,
+      "loss": 3.9015,
+      "step": 1828352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0056058093651382e-05,
+      "loss": 3.8978,
+      "step": 1828864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0047672146140862e-05,
+      "loss": 3.8857,
+      "step": 1829376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.003928619863034e-05,
+      "loss": 3.887,
+      "step": 1829888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.003090025111982e-05,
+      "loss": 3.8916,
+      "step": 1830400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.002253068241303e-05,
+      "loss": 3.8855,
+      "step": 1830912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.001414473490251e-05,
+      "loss": 3.8999,
+      "step": 1831424
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9943506717681885,
+      "eval_runtime": 303.1277,
+      "eval_samples_per_second": 1258.846,
+      "eval_steps_per_second": 39.34,
+      "step": 1831664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.000575878739199e-05,
+      "loss": 3.8962,
+      "step": 1831936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.999737283988147e-05,
+      "loss": 3.8928,
+      "step": 1832448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.998898689237095e-05,
+      "loss": 3.8963,
+      "step": 1832960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.998060094486043e-05,
+      "loss": 3.8955,
+      "step": 1833472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.997221499734991e-05,
+      "loss": 3.9074,
+      "step": 1833984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.996384542864312e-05,
+      "loss": 3.8953,
+      "step": 1834496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.99554594811326e-05,
+      "loss": 3.9005,
+      "step": 1835008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9947073533622083e-05,
+      "loss": 3.8856,
+      "step": 1835520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9938687586111563e-05,
+      "loss": 3.8867,
+      "step": 1836032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9930301638601043e-05,
+      "loss": 3.9046,
+      "step": 1836544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.992191569109052e-05,
+      "loss": 3.8934,
+      "step": 1837056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.991356250118746e-05,
+      "loss": 3.8918,
+      "step": 1837568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9905192932480677e-05,
+      "loss": 3.8999,
+      "step": 1838080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9896806984970157e-05,
+      "loss": 3.8851,
+      "step": 1838592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9888421037459634e-05,
+      "loss": 3.8861,
+      "step": 1839104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9880035089949114e-05,
+      "loss": 3.8823,
+      "step": 1839616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9871649142438594e-05,
+      "loss": 3.8911,
+      "step": 1840128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9863263194928074e-05,
+      "loss": 3.8955,
+      "step": 1840640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9854877247417554e-05,
+      "loss": 3.889,
+      "step": 1841152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9846491299907037e-05,
+      "loss": 3.9027,
+      "step": 1841664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9838105352396517e-05,
+      "loss": 3.9019,
+      "step": 1842176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9829719404885994e-05,
+      "loss": 3.9012,
+      "step": 1842688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9821333457375474e-05,
+      "loss": 3.8921,
+      "step": 1843200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9812947509864954e-05,
+      "loss": 3.8933,
+      "step": 1843712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9804561562354434e-05,
+      "loss": 3.8949,
+      "step": 1844224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9796175614843913e-05,
+      "loss": 3.899,
+      "step": 1844736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9787789667333397e-05,
+      "loss": 3.889,
+      "step": 1845248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9779403719822873e-05,
+      "loss": 3.8928,
+      "step": 1845760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9771034151116086e-05,
+      "loss": 3.8831,
+      "step": 1846272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9762664582409295e-05,
+      "loss": 3.8953,
+      "step": 1846784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9754278634898775e-05,
+      "loss": 3.8897,
+      "step": 1847296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9745892687388258e-05,
+      "loss": 3.9005,
+      "step": 1847808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9737506739877738e-05,
+      "loss": 3.9014,
+      "step": 1848320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9729137171170947e-05,
+      "loss": 3.9036,
+      "step": 1848832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9720751223660427e-05,
+      "loss": 3.8915,
+      "step": 1849344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9712381654953636e-05,
+      "loss": 3.8972,
+      "step": 1849856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.970399570744312e-05,
+      "loss": 3.9008,
+      "step": 1850368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.96956097599326e-05,
+      "loss": 3.8823,
+      "step": 1850880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.968722381242208e-05,
+      "loss": 3.8862,
+      "step": 1851392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.967883786491156e-05,
+      "loss": 3.8827,
+      "step": 1851904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.967045191740104e-05,
+      "loss": 3.8858,
+      "step": 1852416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.966206596989052e-05,
+      "loss": 3.8928,
+      "step": 1852928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9653680022379996e-05,
+      "loss": 3.8972,
+      "step": 1853440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9645294074869476e-05,
+      "loss": 3.8913,
+      "step": 1853952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.963690812735896e-05,
+      "loss": 3.8905,
+      "step": 1854464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.962852217984844e-05,
+      "loss": 3.8906,
+      "step": 1854976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.962013623233792e-05,
+      "loss": 3.8984,
+      "step": 1855488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.961176666363113e-05,
+      "loss": 3.8819,
+      "step": 1856000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.960338071612061e-05,
+      "loss": 3.8691,
+      "step": 1856512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9594994768610088e-05,
+      "loss": 3.9059,
+      "step": 1857024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9586608821099568e-05,
+      "loss": 3.8968,
+      "step": 1857536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.957823925239278e-05,
+      "loss": 3.9008,
+      "step": 1858048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.956985330488226e-05,
+      "loss": 3.8919,
+      "step": 1858560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.956148373617547e-05,
+      "loss": 3.8931,
+      "step": 1859072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.955309778866495e-05,
+      "loss": 3.8861,
+      "step": 1859584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.954471184115443e-05,
+      "loss": 3.8864,
+      "step": 1860096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9536325893643913e-05,
+      "loss": 3.8964,
+      "step": 1860608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9527939946133393e-05,
+      "loss": 3.8805,
+      "step": 1861120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9519553998622873e-05,
+      "loss": 3.9129,
+      "step": 1861632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9511184429916082e-05,
+      "loss": 3.8885,
+      "step": 1862144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9502798482405562e-05,
+      "loss": 3.8738,
+      "step": 1862656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9494412534895042e-05,
+      "loss": 3.9007,
+      "step": 1863168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9486026587384522e-05,
+      "loss": 3.8827,
+      "step": 1863680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9477640639874002e-05,
+      "loss": 3.8731,
+      "step": 1864192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9469254692363482e-05,
+      "loss": 3.9022,
+      "step": 1864704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9460868744852962e-05,
+      "loss": 3.8967,
+      "step": 1865216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9452482797342442e-05,
+      "loss": 3.8835,
+      "step": 1865728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.944411322863565e-05,
+      "loss": 3.8881,
+      "step": 1866240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.943572728112513e-05,
+      "loss": 3.8722,
+      "step": 1866752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9427341333614614e-05,
+      "loss": 3.8857,
+      "step": 1867264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9418955386104094e-05,
+      "loss": 3.894,
+      "step": 1867776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9410569438593574e-05,
+      "loss": 3.8852,
+      "step": 1868288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9402183491083054e-05,
+      "loss": 3.8936,
+      "step": 1868800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9393797543572534e-05,
+      "loss": 3.8869,
+      "step": 1869312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.938541159606201e-05,
+      "loss": 3.9029,
+      "step": 1869824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9377042027355223e-05,
+      "loss": 3.8859,
+      "step": 1870336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9368656079844706e-05,
+      "loss": 3.8929,
+      "step": 1870848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9360270132334183e-05,
+      "loss": 3.8892,
+      "step": 1871360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9351884184823663e-05,
+      "loss": 3.8867,
+      "step": 1871872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9343514616116875e-05,
+      "loss": 3.8919,
+      "step": 1872384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9335145047410084e-05,
+      "loss": 3.8898,
+      "step": 1872896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9326759099899568e-05,
+      "loss": 3.8887,
+      "step": 1873408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9318373152389048e-05,
+      "loss": 3.9004,
+      "step": 1873920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9310003583682257e-05,
+      "loss": 3.8756,
+      "step": 1874432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9301617636171737e-05,
+      "loss": 3.8808,
+      "step": 1874944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9293231688661217e-05,
+      "loss": 3.8951,
+      "step": 1875456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9284845741150697e-05,
+      "loss": 3.8933,
+      "step": 1875968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9276459793640177e-05,
+      "loss": 3.8838,
+      "step": 1876480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9268073846129657e-05,
+      "loss": 3.8963,
+      "step": 1876992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9259687898619137e-05,
+      "loss": 3.8786,
+      "step": 1877504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9251301951108617e-05,
+      "loss": 3.8835,
+      "step": 1878016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.924293238240183e-05,
+      "loss": 3.8813,
+      "step": 1878528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9234546434891306e-05,
+      "loss": 3.8827,
+      "step": 1879040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9226160487380786e-05,
+      "loss": 3.9065,
+      "step": 1879552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.921777453987027e-05,
+      "loss": 3.8894,
+      "step": 1880064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.920938859235975e-05,
+      "loss": 3.8936,
+      "step": 1880576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.920100264484923e-05,
+      "loss": 3.8847,
+      "step": 1881088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.919261669733871e-05,
+      "loss": 3.8858,
+      "step": 1881600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9184247128631918e-05,
+      "loss": 3.8861,
+      "step": 1882112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9175861181121398e-05,
+      "loss": 3.8953,
+      "step": 1882624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9167475233610878e-05,
+      "loss": 3.8899,
+      "step": 1883136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.915910566490409e-05,
+      "loss": 3.8961,
+      "step": 1883648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.915071971739357e-05,
+      "loss": 3.8909,
+      "step": 1884160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.914233376988305e-05,
+      "loss": 3.8893,
+      "step": 1884672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.913394782237253e-05,
+      "loss": 3.8874,
+      "step": 1885184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.912556187486201e-05,
+      "loss": 3.888,
+      "step": 1885696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.911717592735149e-05,
+      "loss": 3.8768,
+      "step": 1886208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9108806358644702e-05,
+      "loss": 3.897,
+      "step": 1886720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9100420411134182e-05,
+      "loss": 3.8793,
+      "step": 1887232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.909203446362366e-05,
+      "loss": 3.889,
+      "step": 1887744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.908364851611314e-05,
+      "loss": 3.8807,
+      "step": 1888256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.907526256860262e-05,
+      "loss": 3.8855,
+      "step": 1888768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.90668766210921e-05,
+      "loss": 3.8873,
+      "step": 1889280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.905849067358158e-05,
+      "loss": 3.8812,
+      "step": 1889792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9050104726071062e-05,
+      "loss": 3.8993,
+      "step": 1890304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9041718778560542e-05,
+      "loss": 3.9043,
+      "step": 1890816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.903334920985375e-05,
+      "loss": 3.8979,
+      "step": 1891328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.902496326234323e-05,
+      "loss": 3.8889,
+      "step": 1891840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.901657731483271e-05,
+      "loss": 3.8823,
+      "step": 1892352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.900819136732219e-05,
+      "loss": 3.884,
+      "step": 1892864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.899980541981167e-05,
+      "loss": 3.8914,
+      "step": 1893376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.899141947230115e-05,
+      "loss": 3.8917,
+      "step": 1893888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.898303352479063e-05,
+      "loss": 3.8798,
+      "step": 1894400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.897464757728011e-05,
+      "loss": 3.8877,
+      "step": 1894912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.896627800857332e-05,
+      "loss": 3.8924,
+      "step": 1895424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.89578920610628e-05,
+      "loss": 3.8905,
+      "step": 1895936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8949522492356016e-05,
+      "loss": 3.8854,
+      "step": 1896448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8941136544845492e-05,
+      "loss": 3.8917,
+      "step": 1896960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8932766976138705e-05,
+      "loss": 3.8836,
+      "step": 1897472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8924381028628185e-05,
+      "loss": 3.8977,
+      "step": 1897984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8915995081117665e-05,
+      "loss": 3.8917,
+      "step": 1898496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8907609133607145e-05,
+      "loss": 3.8994,
+      "step": 1899008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8899223186096625e-05,
+      "loss": 3.8816,
+      "step": 1899520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8890837238586105e-05,
+      "loss": 3.8929,
+      "step": 1900032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8882451291075585e-05,
+      "loss": 3.9,
+      "step": 1900544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8874081722368794e-05,
+      "loss": 3.8898,
+      "step": 1901056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8865695774858274e-05,
+      "loss": 3.8917,
+      "step": 1901568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8857309827347754e-05,
+      "loss": 3.8898,
+      "step": 1902080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8848923879837234e-05,
+      "loss": 3.8925,
+      "step": 1902592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8840537932326717e-05,
+      "loss": 3.9092,
+      "step": 1903104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8832151984816197e-05,
+      "loss": 3.8857,
+      "step": 1903616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8823766037305673e-05,
+      "loss": 3.8944,
+      "step": 1904128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8815380089795153e-05,
+      "loss": 3.8928,
+      "step": 1904640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8806994142284633e-05,
+      "loss": 3.8913,
+      "step": 1905152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8798624573577846e-05,
+      "loss": 3.8829,
+      "step": 1905664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8790238626067326e-05,
+      "loss": 3.8804,
+      "step": 1906176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8781852678556806e-05,
+      "loss": 3.8878,
+      "step": 1906688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8773466731046286e-05,
+      "loss": 3.8839,
+      "step": 1907200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8765080783535766e-05,
+      "loss": 3.8906,
+      "step": 1907712
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9927806854248047,
+      "eval_runtime": 298.797,
+      "eval_samples_per_second": 1277.091,
+      "eval_steps_per_second": 39.91,
+      "step": 1907984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8756711214828975e-05,
+      "loss": 3.8926,
+      "step": 1908224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8748325267318455e-05,
+      "loss": 3.8882,
+      "step": 1908736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8739939319807938e-05,
+      "loss": 3.8884,
+      "step": 1909248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8731553372297418e-05,
+      "loss": 3.8917,
+      "step": 1909760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8723167424786898e-05,
+      "loss": 3.9032,
+      "step": 1910272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8714781477276378e-05,
+      "loss": 3.8931,
+      "step": 1910784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8706395529765855e-05,
+      "loss": 3.8916,
+      "step": 1911296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8698009582255334e-05,
+      "loss": 3.8792,
+      "step": 1911808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8689623634744814e-05,
+      "loss": 3.881,
+      "step": 1912320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8681237687234294e-05,
+      "loss": 3.8978,
+      "step": 1912832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8672851739723778e-05,
+      "loss": 3.8885,
+      "step": 1913344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8664465792213258e-05,
+      "loss": 3.8879,
+      "step": 1913856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8656096223506467e-05,
+      "loss": 3.8938,
+      "step": 1914368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8647710275995947e-05,
+      "loss": 3.8794,
+      "step": 1914880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8639324328485427e-05,
+      "loss": 3.8841,
+      "step": 1915392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8630938380974907e-05,
+      "loss": 3.8708,
+      "step": 1915904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8622552433464387e-05,
+      "loss": 3.8872,
+      "step": 1916416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8614166485953867e-05,
+      "loss": 3.8921,
+      "step": 1916928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8605780538443347e-05,
+      "loss": 3.881,
+      "step": 1917440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8597394590932826e-05,
+      "loss": 3.8976,
+      "step": 1917952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8589008643422306e-05,
+      "loss": 3.8961,
+      "step": 1918464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8580639074715516e-05,
+      "loss": 3.8951,
+      "step": 1918976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8572253127204995e-05,
+      "loss": 3.8901,
+      "step": 1919488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.856386717969448e-05,
+      "loss": 3.8912,
+      "step": 1920000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.855548123218396e-05,
+      "loss": 3.8922,
+      "step": 1920512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.854709528467344e-05,
+      "loss": 3.8894,
+      "step": 1921024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8538725715966648e-05,
+      "loss": 3.8813,
+      "step": 1921536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8530339768456128e-05,
+      "loss": 3.8891,
+      "step": 1922048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8521953820945608e-05,
+      "loss": 3.8824,
+      "step": 1922560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8513567873435088e-05,
+      "loss": 3.8879,
+      "step": 1923072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.85051983047283e-05,
+      "loss": 3.8904,
+      "step": 1923584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.849681235721778e-05,
+      "loss": 3.8901,
+      "step": 1924096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.848842640970726e-05,
+      "loss": 3.8943,
+      "step": 1924608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.848004046219674e-05,
+      "loss": 3.9014,
+      "step": 1925120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8471654514686217e-05,
+      "loss": 3.8853,
+      "step": 1925632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8463284945979432e-05,
+      "loss": 3.889,
+      "step": 1926144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8454898998468912e-05,
+      "loss": 3.8984,
+      "step": 1926656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.844651305095839e-05,
+      "loss": 3.8775,
+      "step": 1927168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.843812710344787e-05,
+      "loss": 3.8816,
+      "step": 1927680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.842975753474108e-05,
+      "loss": 3.8784,
+      "step": 1928192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.842137158723056e-05,
+      "loss": 3.8786,
+      "step": 1928704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.841298563972004e-05,
+      "loss": 3.8878,
+      "step": 1929216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.840459969220952e-05,
+      "loss": 3.8953,
+      "step": 1929728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8396213744699e-05,
+      "loss": 3.8827,
+      "step": 1930240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.838782779718848e-05,
+      "loss": 3.8862,
+      "step": 1930752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.837944184967796e-05,
+      "loss": 3.8822,
+      "step": 1931264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.837105590216744e-05,
+      "loss": 3.8943,
+      "step": 1931776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.836266995465692e-05,
+      "loss": 3.8764,
+      "step": 1932288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.83542840071464e-05,
+      "loss": 3.8667,
+      "step": 1932800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.834589805963588e-05,
+      "loss": 3.8987,
+      "step": 1933312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.833751211212536e-05,
+      "loss": 3.8917,
+      "step": 1933824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.832914254341857e-05,
+      "loss": 3.8994,
+      "step": 1934336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.832075659590805e-05,
+      "loss": 3.8805,
+      "step": 1934848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.831237064839753e-05,
+      "loss": 3.885,
+      "step": 1935360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.830398470088701e-05,
+      "loss": 3.8825,
+      "step": 1935872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8295631510983955e-05,
+      "loss": 3.8805,
+      "step": 1936384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8287245563473435e-05,
+      "loss": 3.8925,
+      "step": 1936896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8278859615962915e-05,
+      "loss": 3.8722,
+      "step": 1937408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8270473668452395e-05,
+      "loss": 3.9073,
+      "step": 1937920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.826208772094187e-05,
+      "loss": 3.8869,
+      "step": 1938432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8253701773431355e-05,
+      "loss": 3.8629,
+      "step": 1938944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8245315825920835e-05,
+      "loss": 3.8968,
+      "step": 1939456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8236946257214044e-05,
+      "loss": 3.8798,
+      "step": 1939968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8228560309703524e-05,
+      "loss": 3.8648,
+      "step": 1940480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8220174362193004e-05,
+      "loss": 3.8942,
+      "step": 1940992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8211788414682484e-05,
+      "loss": 3.8929,
+      "step": 1941504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8203402467171964e-05,
+      "loss": 3.8788,
+      "step": 1942016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8195016519661447e-05,
+      "loss": 3.885,
+      "step": 1942528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8186630572150927e-05,
+      "loss": 3.8657,
+      "step": 1943040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8178244624640403e-05,
+      "loss": 3.88,
+      "step": 1943552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8169875055933616e-05,
+      "loss": 3.8851,
+      "step": 1944064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8161489108423096e-05,
+      "loss": 3.8816,
+      "step": 1944576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8153103160912572e-05,
+      "loss": 3.8896,
+      "step": 1945088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8144717213402056e-05,
+      "loss": 3.8787,
+      "step": 1945600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8136347644695268e-05,
+      "loss": 3.9008,
+      "step": 1946112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8127961697184748e-05,
+      "loss": 3.8861,
+      "step": 1946624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8119575749674225e-05,
+      "loss": 3.8838,
+      "step": 1947136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8111206180967437e-05,
+      "loss": 3.8843,
+      "step": 1947648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8102820233456917e-05,
+      "loss": 3.8829,
+      "step": 1948160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8094434285946397e-05,
+      "loss": 3.8856,
+      "step": 1948672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8086048338435877e-05,
+      "loss": 3.885,
+      "step": 1949184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8077662390925357e-05,
+      "loss": 3.8852,
+      "step": 1949696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.806929282221857e-05,
+      "loss": 3.8946,
+      "step": 1950208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8060906874708046e-05,
+      "loss": 3.8731,
+      "step": 1950720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8052520927197526e-05,
+      "loss": 3.8731,
+      "step": 1951232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.804413497968701e-05,
+      "loss": 3.8864,
+      "step": 1951744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.803574903217649e-05,
+      "loss": 3.8927,
+      "step": 1952256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.802736308466597e-05,
+      "loss": 3.876,
+      "step": 1952768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.801897713715545e-05,
+      "loss": 3.8889,
+      "step": 1953280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.801059118964493e-05,
+      "loss": 3.8752,
+      "step": 1953792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.800222162093814e-05,
+      "loss": 3.8762,
+      "step": 1954304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.799383567342762e-05,
+      "loss": 3.8778,
+      "step": 1954816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.79854497259171e-05,
+      "loss": 3.8819,
+      "step": 1955328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.797706377840658e-05,
+      "loss": 3.8962,
+      "step": 1955840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.796869420969979e-05,
+      "loss": 3.8857,
+      "step": 1956352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.796030826218927e-05,
+      "loss": 3.8892,
+      "step": 1956864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.795192231467875e-05,
+      "loss": 3.8796,
+      "step": 1957376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7943536367168227e-05,
+      "loss": 3.8779,
+      "step": 1957888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.793515041965771e-05,
+      "loss": 3.8824,
+      "step": 1958400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.792676447214719e-05,
+      "loss": 3.8885,
+      "step": 1958912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.791837852463667e-05,
+      "loss": 3.8875,
+      "step": 1959424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.790999257712615e-05,
+      "loss": 3.8906,
+      "step": 1959936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.790162300841936e-05,
+      "loss": 3.8877,
+      "step": 1960448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.789323706090884e-05,
+      "loss": 3.878,
+      "step": 1960960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.788485111339832e-05,
+      "loss": 3.8863,
+      "step": 1961472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7876465165887803e-05,
+      "loss": 3.8823,
+      "step": 1961984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7868079218377283e-05,
+      "loss": 3.8759,
+      "step": 1962496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7859709649670492e-05,
+      "loss": 3.8839,
+      "step": 1963008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7851323702159972e-05,
+      "loss": 3.8778,
+      "step": 1963520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7842937754649452e-05,
+      "loss": 3.881,
+      "step": 1964032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7834551807138932e-05,
+      "loss": 3.8827,
+      "step": 1964544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.782616585962841e-05,
+      "loss": 3.8765,
+      "step": 1965056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.781777991211789e-05,
+      "loss": 3.8826,
+      "step": 1965568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.780939396460737e-05,
+      "loss": 3.8773,
+      "step": 1966080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.780100801709685e-05,
+      "loss": 3.8942,
+      "step": 1966592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.779263844839006e-05,
+      "loss": 3.8949,
+      "step": 1967104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7784268879683273e-05,
+      "loss": 3.8927,
+      "step": 1967616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7775882932172756e-05,
+      "loss": 3.8856,
+      "step": 1968128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7767496984662236e-05,
+      "loss": 3.8769,
+      "step": 1968640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7759111037151713e-05,
+      "loss": 3.8791,
+      "step": 1969152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7750725089641193e-05,
+      "loss": 3.8861,
+      "step": 1969664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7742355520934405e-05,
+      "loss": 3.8889,
+      "step": 1970176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7733969573423882e-05,
+      "loss": 3.8713,
+      "step": 1970688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7725583625913365e-05,
+      "loss": 3.8835,
+      "step": 1971200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7717197678402845e-05,
+      "loss": 3.8851,
+      "step": 1971712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7708811730892325e-05,
+      "loss": 3.8879,
+      "step": 1972224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7700442162185534e-05,
+      "loss": 3.8792,
+      "step": 1972736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7692056214675014e-05,
+      "loss": 3.8839,
+      "step": 1973248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7683670267164494e-05,
+      "loss": 3.8771,
+      "step": 1973760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.767530069845771e-05,
+      "loss": 3.8958,
+      "step": 1974272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7666914750947187e-05,
+      "loss": 3.8856,
+      "step": 1974784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7658528803436667e-05,
+      "loss": 3.8939,
+      "step": 1975296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7650142855926147e-05,
+      "loss": 3.8778,
+      "step": 1975808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7641756908415627e-05,
+      "loss": 3.8886,
+      "step": 1976320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7633370960905107e-05,
+      "loss": 3.8928,
+      "step": 1976832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7624985013394586e-05,
+      "loss": 3.8817,
+      "step": 1977344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7616599065884066e-05,
+      "loss": 3.8869,
+      "step": 1977856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7608213118373546e-05,
+      "loss": 3.8864,
+      "step": 1978368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.759984354966676e-05,
+      "loss": 3.8877,
+      "step": 1978880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.759145760215624e-05,
+      "loss": 3.8988,
+      "step": 1979392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7583071654645715e-05,
+      "loss": 3.8837,
+      "step": 1979904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7574685707135195e-05,
+      "loss": 3.8892,
+      "step": 1980416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.756629975962468e-05,
+      "loss": 3.8884,
+      "step": 1980928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.755791381211416e-05,
+      "loss": 3.8876,
+      "step": 1981440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7549544243407368e-05,
+      "loss": 3.8776,
+      "step": 1981952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7541158295896848e-05,
+      "loss": 3.8757,
+      "step": 1982464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7532772348386328e-05,
+      "loss": 3.8795,
+      "step": 1982976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7524386400875808e-05,
+      "loss": 3.8819,
+      "step": 1983488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7516000453365288e-05,
+      "loss": 3.8829,
+      "step": 1984000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.991046905517578,
+      "eval_runtime": 296.5186,
+      "eval_samples_per_second": 1286.904,
+      "eval_steps_per_second": 40.217,
+      "step": 1984304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7507614505854768e-05,
+      "loss": 3.8839,
+      "step": 1984512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7499228558344247e-05,
+      "loss": 3.8808,
+      "step": 1985024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7490842610833727e-05,
+      "loss": 3.8853,
+      "step": 1985536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.748247304212694e-05,
+      "loss": 3.8883,
+      "step": 1986048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.747408709461642e-05,
+      "loss": 3.8969,
+      "step": 1986560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7465701147105896e-05,
+      "loss": 3.8858,
+      "step": 1987072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.745731519959538e-05,
+      "loss": 3.8869,
+      "step": 1987584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.744892925208486e-05,
+      "loss": 3.8786,
+      "step": 1988096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.744055968337807e-05,
+      "loss": 3.8752,
+      "step": 1988608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.743217373586755e-05,
+      "loss": 3.8874,
+      "step": 1989120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.742378778835703e-05,
+      "loss": 3.8908,
+      "step": 1989632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.741540184084651e-05,
+      "loss": 3.8798,
+      "step": 1990144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.740703227213972e-05,
+      "loss": 3.8895,
+      "step": 1990656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.73986463246292e-05,
+      "loss": 3.873,
+      "step": 1991168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7390276755922414e-05,
+      "loss": 3.8792,
+      "step": 1991680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7381890808411894e-05,
+      "loss": 3.8627,
+      "step": 1992192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.737350486090137e-05,
+      "loss": 3.8864,
+      "step": 1992704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.736511891339085e-05,
+      "loss": 3.8873,
+      "step": 1993216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7356749344684066e-05,
+      "loss": 3.8766,
+      "step": 1993728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7348363397173543e-05,
+      "loss": 3.888,
+      "step": 1994240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7339977449663022e-05,
+      "loss": 3.893,
+      "step": 1994752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7331591502152502e-05,
+      "loss": 3.8884,
+      "step": 1995264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7323205554641982e-05,
+      "loss": 3.8854,
+      "step": 1995776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7314835985935195e-05,
+      "loss": 3.8834,
+      "step": 1996288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7306450038424675e-05,
+      "loss": 3.8865,
+      "step": 1996800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7298064090914155e-05,
+      "loss": 3.8912,
+      "step": 1997312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7289678143403635e-05,
+      "loss": 3.8726,
+      "step": 1997824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7281292195893115e-05,
+      "loss": 3.8854,
+      "step": 1998336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7272906248382595e-05,
+      "loss": 3.875,
+      "step": 1998848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7264520300872075e-05,
+      "loss": 3.8791,
+      "step": 1999360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.725613435336155e-05,
+      "loss": 3.886,
+      "step": 1999872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7247764784654767e-05,
+      "loss": 3.887,
+      "step": 2000384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7239378837144247e-05,
+      "loss": 3.8866,
+      "step": 2000896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7231009268437456e-05,
+      "loss": 3.8967,
+      "step": 2001408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7222623320926936e-05,
+      "loss": 3.8824,
+      "step": 2001920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.721425375222015e-05,
+      "loss": 3.8858,
+      "step": 2002432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.720586780470963e-05,
+      "loss": 3.892,
+      "step": 2002944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.719748185719911e-05,
+      "loss": 3.8717,
+      "step": 2003456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.718909590968859e-05,
+      "loss": 3.875,
+      "step": 2003968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.718070996217807e-05,
+      "loss": 3.8785,
+      "step": 2004480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.717232401466755e-05,
+      "loss": 3.8736,
+      "step": 2004992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7163938067157025e-05,
+      "loss": 3.8806,
+      "step": 2005504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7155552119646505e-05,
+      "loss": 3.8916,
+      "step": 2006016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7147166172135988e-05,
+      "loss": 3.8773,
+      "step": 2006528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7138796603429197e-05,
+      "loss": 3.8801,
+      "step": 2007040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7130410655918677e-05,
+      "loss": 3.878,
+      "step": 2007552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7122024708408157e-05,
+      "loss": 3.8881,
+      "step": 2008064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7113638760897637e-05,
+      "loss": 3.8713,
+      "step": 2008576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7105252813387117e-05,
+      "loss": 3.8616,
+      "step": 2009088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.709688324468033e-05,
+      "loss": 3.8914,
+      "step": 2009600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.708849729716981e-05,
+      "loss": 3.8857,
+      "step": 2010112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.708011134965929e-05,
+      "loss": 3.8953,
+      "step": 2010624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.707172540214877e-05,
+      "loss": 3.8764,
+      "step": 2011136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.706335583344198e-05,
+      "loss": 3.88,
+      "step": 2011648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.705496988593146e-05,
+      "loss": 3.8762,
+      "step": 2012160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7046583938420942e-05,
+      "loss": 3.8743,
+      "step": 2012672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7038197990910422e-05,
+      "loss": 3.8908,
+      "step": 2013184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7029812043399902e-05,
+      "loss": 3.8636,
+      "step": 2013696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.702142609588938e-05,
+      "loss": 3.9055,
+      "step": 2014208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.701305652718259e-05,
+      "loss": 3.8855,
+      "step": 2014720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.700467057967207e-05,
+      "loss": 3.8584,
+      "step": 2015232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.699628463216155e-05,
+      "loss": 3.8886,
+      "step": 2015744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.698789868465103e-05,
+      "loss": 3.8763,
+      "step": 2016256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.697951273714051e-05,
+      "loss": 3.8628,
+      "step": 2016768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.697112678962999e-05,
+      "loss": 3.8865,
+      "step": 2017280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.696274084211947e-05,
+      "loss": 3.8899,
+      "step": 2017792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.695435489460895e-05,
+      "loss": 3.8718,
+      "step": 2018304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.694598532590216e-05,
+      "loss": 3.8831,
+      "step": 2018816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6937615757195375e-05,
+      "loss": 3.86,
+      "step": 2019328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6929229809684852e-05,
+      "loss": 3.8732,
+      "step": 2019840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6920843862174332e-05,
+      "loss": 3.8809,
+      "step": 2020352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6912457914663812e-05,
+      "loss": 3.8751,
+      "step": 2020864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6904071967153292e-05,
+      "loss": 3.8845,
+      "step": 2021376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6895686019642772e-05,
+      "loss": 3.8751,
+      "step": 2021888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6887316450935984e-05,
+      "loss": 3.8951,
+      "step": 2022400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6878930503425464e-05,
+      "loss": 3.876,
+      "step": 2022912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6870544555914944e-05,
+      "loss": 3.8853,
+      "step": 2023424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6862158608404424e-05,
+      "loss": 3.8771,
+      "step": 2023936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6853772660893904e-05,
+      "loss": 3.8761,
+      "step": 2024448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6845403092187113e-05,
+      "loss": 3.8803,
+      "step": 2024960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6837017144676597e-05,
+      "loss": 3.879,
+      "step": 2025472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6828631197166077e-05,
+      "loss": 3.8827,
+      "step": 2025984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6820245249655557e-05,
+      "loss": 3.8897,
+      "step": 2026496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6811875680948766e-05,
+      "loss": 3.8737,
+      "step": 2027008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6803506112241975e-05,
+      "loss": 3.863,
+      "step": 2027520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6795120164731458e-05,
+      "loss": 3.8794,
+      "step": 2028032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6786734217220938e-05,
+      "loss": 3.8865,
+      "step": 2028544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6778348269710418e-05,
+      "loss": 3.873,
+      "step": 2029056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6769962322199898e-05,
+      "loss": 3.8841,
+      "step": 2029568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6761576374689378e-05,
+      "loss": 3.8717,
+      "step": 2030080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6753190427178854e-05,
+      "loss": 3.8715,
+      "step": 2030592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6744804479668334e-05,
+      "loss": 3.8734,
+      "step": 2031104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6736418532157814e-05,
+      "loss": 3.8748,
+      "step": 2031616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.672804896345103e-05,
+      "loss": 3.8923,
+      "step": 2032128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6719663015940507e-05,
+      "loss": 3.8821,
+      "step": 2032640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6711277068429987e-05,
+      "loss": 3.8825,
+      "step": 2033152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6702891120919467e-05,
+      "loss": 3.8769,
+      "step": 2033664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6694505173408947e-05,
+      "loss": 3.8724,
+      "step": 2034176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6686119225898427e-05,
+      "loss": 3.876,
+      "step": 2034688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6677733278387907e-05,
+      "loss": 3.8825,
+      "step": 2035200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.666934733087739e-05,
+      "loss": 3.8865,
+      "step": 2035712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.66609777621706e-05,
+      "loss": 3.8825,
+      "step": 2036224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.665259181466008e-05,
+      "loss": 3.884,
+      "step": 2036736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.664420586714956e-05,
+      "loss": 3.8728,
+      "step": 2037248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6635836298442768e-05,
+      "loss": 3.8824,
+      "step": 2037760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.662745035093225e-05,
+      "loss": 3.8771,
+      "step": 2038272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.661906440342173e-05,
+      "loss": 3.8714,
+      "step": 2038784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.661067845591121e-05,
+      "loss": 3.8792,
+      "step": 2039296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6602292508400688e-05,
+      "loss": 3.8738,
+      "step": 2039808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6593906560890168e-05,
+      "loss": 3.881,
+      "step": 2040320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6585520613379648e-05,
+      "loss": 3.8702,
+      "step": 2040832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6577134665869128e-05,
+      "loss": 3.8731,
+      "step": 2041344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.656876509716234e-05,
+      "loss": 3.8799,
+      "step": 2041856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.656037914965182e-05,
+      "loss": 3.8769,
+      "step": 2042368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.65519932021413e-05,
+      "loss": 3.8845,
+      "step": 2042880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.654360725463078e-05,
+      "loss": 3.8876,
+      "step": 2043392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.653523768592399e-05,
+      "loss": 3.8897,
+      "step": 2043904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.652685173841347e-05,
+      "loss": 3.8799,
+      "step": 2044416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6518465790902952e-05,
+      "loss": 3.8751,
+      "step": 2044928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6510079843392432e-05,
+      "loss": 3.8736,
+      "step": 2045440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6501693895881912e-05,
+      "loss": 3.8882,
+      "step": 2045952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6493340705978854e-05,
+      "loss": 3.8804,
+      "step": 2046464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6484954758468334e-05,
+      "loss": 3.8704,
+      "step": 2046976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6476568810957814e-05,
+      "loss": 3.8735,
+      "step": 2047488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6468182863447294e-05,
+      "loss": 3.8838,
+      "step": 2048000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6459796915936774e-05,
+      "loss": 3.8809,
+      "step": 2048512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6451427347229983e-05,
+      "loss": 3.8754,
+      "step": 2049024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6443041399719463e-05,
+      "loss": 3.8777,
+      "step": 2049536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6434655452208943e-05,
+      "loss": 3.8735,
+      "step": 2050048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6426269504698423e-05,
+      "loss": 3.8876,
+      "step": 2050560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6417883557187906e-05,
+      "loss": 3.8834,
+      "step": 2051072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6409513988481115e-05,
+      "loss": 3.8865,
+      "step": 2051584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6401128040970595e-05,
+      "loss": 3.8759,
+      "step": 2052096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6392742093460075e-05,
+      "loss": 3.8852,
+      "step": 2052608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6384356145949555e-05,
+      "loss": 3.8874,
+      "step": 2053120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6375986577242768e-05,
+      "loss": 3.8729,
+      "step": 2053632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6367600629732248e-05,
+      "loss": 3.8816,
+      "step": 2054144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6359214682221728e-05,
+      "loss": 3.8834,
+      "step": 2054656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6350828734711207e-05,
+      "loss": 3.8818,
+      "step": 2055168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6342459166004417e-05,
+      "loss": 3.8947,
+      "step": 2055680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6334073218493897e-05,
+      "loss": 3.8811,
+      "step": 2056192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6325687270983376e-05,
+      "loss": 3.8814,
+      "step": 2056704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.631730132347286e-05,
+      "loss": 3.8867,
+      "step": 2057216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.630891537596234e-05,
+      "loss": 3.8789,
+      "step": 2057728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6300529428451816e-05,
+      "loss": 3.8742,
+      "step": 2058240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.629215985974503e-05,
+      "loss": 3.8728,
+      "step": 2058752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.628377391223451e-05,
+      "loss": 3.8718,
+      "step": 2059264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.627538796472399e-05,
+      "loss": 3.8787,
+      "step": 2059776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.626700201721347e-05,
+      "loss": 3.879,
+      "step": 2060288
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.989319324493408,
+      "eval_runtime": 301.7396,
+      "eval_samples_per_second": 1264.637,
+      "eval_steps_per_second": 39.521,
+      "step": 2060624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.625863244850668e-05,
+      "loss": 3.8766,
+      "step": 2060800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.625024650099616e-05,
+      "loss": 3.8765,
+      "step": 2061312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6241860553485638e-05,
+      "loss": 3.8798,
+      "step": 2061824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6233474605975118e-05,
+      "loss": 3.8818,
+      "step": 2062336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6225088658464598e-05,
+      "loss": 3.8944,
+      "step": 2062848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6216702710954078e-05,
+      "loss": 3.8822,
+      "step": 2063360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.620831676344356e-05,
+      "loss": 3.884,
+      "step": 2063872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.619993081593304e-05,
+      "loss": 3.8763,
+      "step": 2064384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.619154486842252e-05,
+      "loss": 3.87,
+      "step": 2064896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6183158920911997e-05,
+      "loss": 3.8784,
+      "step": 2065408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6174772973401477e-05,
+      "loss": 3.8907,
+      "step": 2065920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6166387025890957e-05,
+      "loss": 3.8705,
+      "step": 2066432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.615801745718417e-05,
+      "loss": 3.8875,
+      "step": 2066944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.614963150967365e-05,
+      "loss": 3.8685,
+      "step": 2067456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6141261940966862e-05,
+      "loss": 3.8728,
+      "step": 2067968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6132875993456342e-05,
+      "loss": 3.8621,
+      "step": 2068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.612449004594582e-05,
+      "loss": 3.8808,
+      "step": 2068992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.61161040984353e-05,
+      "loss": 3.8822,
+      "step": 2069504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6107718150924782e-05,
+      "loss": 3.8756,
+      "step": 2070016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6099332203414262e-05,
+      "loss": 3.8817,
+      "step": 2070528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6090946255903742e-05,
+      "loss": 3.8911,
+      "step": 2071040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6082560308393222e-05,
+      "loss": 3.8818,
+      "step": 2071552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6074174360882702e-05,
+      "loss": 3.8787,
+      "step": 2072064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.606578841337218e-05,
+      "loss": 3.8806,
+      "step": 2072576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.605740246586166e-05,
+      "loss": 3.8843,
+      "step": 2073088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.604901651835114e-05,
+      "loss": 3.8835,
+      "step": 2073600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.604063057084062e-05,
+      "loss": 3.8693,
+      "step": 2074112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.603226100213383e-05,
+      "loss": 3.8801,
+      "step": 2074624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.602387505462331e-05,
+      "loss": 3.8721,
+      "step": 2075136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6015505485916523e-05,
+      "loss": 3.8716,
+      "step": 2075648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6007119538406e-05,
+      "loss": 3.882,
+      "step": 2076160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5998733590895483e-05,
+      "loss": 3.8794,
+      "step": 2076672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5990347643384963e-05,
+      "loss": 3.8849,
+      "step": 2077184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5981961695874443e-05,
+      "loss": 3.8894,
+      "step": 2077696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5973575748363923e-05,
+      "loss": 3.8758,
+      "step": 2078208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5965206179657132e-05,
+      "loss": 3.8851,
+      "step": 2078720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5956820232146612e-05,
+      "loss": 3.8846,
+      "step": 2079232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5948434284636092e-05,
+      "loss": 3.8644,
+      "step": 2079744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5940048337125575e-05,
+      "loss": 3.8731,
+      "step": 2080256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5931662389615055e-05,
+      "loss": 3.8681,
+      "step": 2080768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5923292820908264e-05,
+      "loss": 3.8677,
+      "step": 2081280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5914906873397744e-05,
+      "loss": 3.8779,
+      "step": 2081792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5906520925887224e-05,
+      "loss": 3.8895,
+      "step": 2082304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5898134978376704e-05,
+      "loss": 3.8708,
+      "step": 2082816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5889749030866184e-05,
+      "loss": 3.8739,
+      "step": 2083328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5881363083355664e-05,
+      "loss": 3.8777,
+      "step": 2083840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5872977135845144e-05,
+      "loss": 3.8819,
+      "step": 2084352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5864591188334624e-05,
+      "loss": 3.87,
+      "step": 2084864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5856205240824104e-05,
+      "loss": 3.8566,
+      "step": 2085376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5847819293313584e-05,
+      "loss": 3.8836,
+      "step": 2085888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5839433345803064e-05,
+      "loss": 3.8844,
+      "step": 2086400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5831047398292544e-05,
+      "loss": 3.8888,
+      "step": 2086912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5822677829585756e-05,
+      "loss": 3.8754,
+      "step": 2087424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5814291882075236e-05,
+      "loss": 3.8718,
+      "step": 2087936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5805905934564713e-05,
+      "loss": 3.8765,
+      "step": 2088448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5797536365857925e-05,
+      "loss": 3.8642,
+      "step": 2088960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5789150418347405e-05,
+      "loss": 3.8854,
+      "step": 2089472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5780780849640618e-05,
+      "loss": 3.8593,
+      "step": 2089984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.577241128093383e-05,
+      "loss": 3.9002,
+      "step": 2090496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5764025333423307e-05,
+      "loss": 3.8775,
+      "step": 2091008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5755639385912787e-05,
+      "loss": 3.8589,
+      "step": 2091520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5747253438402267e-05,
+      "loss": 3.8817,
+      "step": 2092032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5738867490891747e-05,
+      "loss": 3.8725,
+      "step": 2092544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.573048154338123e-05,
+      "loss": 3.8576,
+      "step": 2093056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.572209559587071e-05,
+      "loss": 3.8802,
+      "step": 2093568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5713709648360187e-05,
+      "loss": 3.8874,
+      "step": 2094080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5705323700849667e-05,
+      "loss": 3.868,
+      "step": 2094592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5696937753339147e-05,
+      "loss": 3.8809,
+      "step": 2095104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5688551805828626e-05,
+      "loss": 3.8538,
+      "step": 2095616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5680165858318106e-05,
+      "loss": 3.8699,
+      "step": 2096128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5671779910807586e-05,
+      "loss": 3.8705,
+      "step": 2096640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.56634103421008e-05,
+      "loss": 3.8778,
+      "step": 2097152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.565502439459028e-05,
+      "loss": 3.8762,
+      "step": 2097664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.564663844707976e-05,
+      "loss": 3.8701,
+      "step": 2098176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.563825249956924e-05,
+      "loss": 3.8922,
+      "step": 2098688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5629866552058715e-05,
+      "loss": 3.8751,
+      "step": 2099200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.56214806045482e-05,
+      "loss": 3.8804,
+      "step": 2099712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.561309465703768e-05,
+      "loss": 3.8704,
+      "step": 2100224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.560470870952716e-05,
+      "loss": 3.8711,
+      "step": 2100736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.55963555196241e-05,
+      "loss": 3.8797,
+      "step": 2101248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.558796957211358e-05,
+      "loss": 3.874,
+      "step": 2101760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.557958362460306e-05,
+      "loss": 3.8735,
+      "step": 2102272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.557119767709254e-05,
+      "loss": 3.8871,
+      "step": 2102784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.556281172958202e-05,
+      "loss": 3.8704,
+      "step": 2103296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5554442160875233e-05,
+      "loss": 3.8615,
+      "step": 2103808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5546056213364712e-05,
+      "loss": 3.8735,
+      "step": 2104320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.553767026585419e-05,
+      "loss": 3.8825,
+      "step": 2104832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.552928431834367e-05,
+      "loss": 3.8679,
+      "step": 2105344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5520898370833152e-05,
+      "loss": 3.8807,
+      "step": 2105856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5512528802126365e-05,
+      "loss": 3.8718,
+      "step": 2106368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.550414285461584e-05,
+      "loss": 3.8624,
+      "step": 2106880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.549575690710532e-05,
+      "loss": 3.8719,
+      "step": 2107392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.54873709595948e-05,
+      "loss": 3.8684,
+      "step": 2107904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5479001390888014e-05,
+      "loss": 3.8862,
+      "step": 2108416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5470615443377494e-05,
+      "loss": 3.8762,
+      "step": 2108928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5462229495866974e-05,
+      "loss": 3.8831,
+      "step": 2109440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5453843548356454e-05,
+      "loss": 3.8694,
+      "step": 2109952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5445457600845934e-05,
+      "loss": 3.8679,
+      "step": 2110464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5437088032139143e-05,
+      "loss": 3.8749,
+      "step": 2110976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5428702084628623e-05,
+      "loss": 3.8786,
+      "step": 2111488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5420316137118103e-05,
+      "loss": 3.8778,
+      "step": 2112000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5411930189607586e-05,
+      "loss": 3.8798,
+      "step": 2112512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5403560620900795e-05,
+      "loss": 3.8774,
+      "step": 2113024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5395174673390275e-05,
+      "loss": 3.8681,
+      "step": 2113536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5386788725879755e-05,
+      "loss": 3.8815,
+      "step": 2114048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5378402778369235e-05,
+      "loss": 3.8726,
+      "step": 2114560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5370016830858715e-05,
+      "loss": 3.8708,
+      "step": 2115072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5361630883348195e-05,
+      "loss": 3.8731,
+      "step": 2115584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5353244935837675e-05,
+      "loss": 3.87,
+      "step": 2116096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5344858988327155e-05,
+      "loss": 3.8718,
+      "step": 2116608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5336505798424096e-05,
+      "loss": 3.8673,
+      "step": 2117120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5328119850913576e-05,
+      "loss": 3.8677,
+      "step": 2117632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5319733903403056e-05,
+      "loss": 3.8754,
+      "step": 2118144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.531134795589254e-05,
+      "loss": 3.8744,
+      "step": 2118656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.530296200838202e-05,
+      "loss": 3.8775,
+      "step": 2119168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5294576060871496e-05,
+      "loss": 3.8808,
+      "step": 2119680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.528620649216471e-05,
+      "loss": 3.8906,
+      "step": 2120192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.527782054465419e-05,
+      "loss": 3.8718,
+      "step": 2120704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.526943459714367e-05,
+      "loss": 3.8699,
+      "step": 2121216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.526104864963315e-05,
+      "loss": 3.8706,
+      "step": 2121728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5252662702122627e-05,
+      "loss": 3.878,
+      "step": 2122240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5244276754612108e-05,
+      "loss": 3.8755,
+      "step": 2122752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5235907185905318e-05,
+      "loss": 3.8704,
+      "step": 2123264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.52275212383948e-05,
+      "loss": 3.8702,
+      "step": 2123776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5219135290884279e-05,
+      "loss": 3.8797,
+      "step": 2124288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5210749343373759e-05,
+      "loss": 3.8761,
+      "step": 2124800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.520237977466697e-05,
+      "loss": 3.8675,
+      "step": 2125312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.519399382715645e-05,
+      "loss": 3.8759,
+      "step": 2125824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.518560787964593e-05,
+      "loss": 3.8724,
+      "step": 2126336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.517722193213541e-05,
+      "loss": 3.8773,
+      "step": 2126848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5168835984624891e-05,
+      "loss": 3.8813,
+      "step": 2127360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.51604664159181e-05,
+      "loss": 3.8835,
+      "step": 2127872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.515208046840758e-05,
+      "loss": 3.8711,
+      "step": 2128384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5143694520897062e-05,
+      "loss": 3.8801,
+      "step": 2128896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5135308573386542e-05,
+      "loss": 3.8816,
+      "step": 2129408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5126939004679753e-05,
+      "loss": 3.8709,
+      "step": 2129920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5118553057169233e-05,
+      "loss": 3.874,
+      "step": 2130432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5110167109658713e-05,
+      "loss": 3.8782,
+      "step": 2130944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5101781162148193e-05,
+      "loss": 3.8863,
+      "step": 2131456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5093395214637673e-05,
+      "loss": 3.8852,
+      "step": 2131968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5085009267127151e-05,
+      "loss": 3.874,
+      "step": 2132480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5076623319616631e-05,
+      "loss": 3.8773,
+      "step": 2132992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.506823737210611e-05,
+      "loss": 3.8806,
+      "step": 2133504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5059867803399322e-05,
+      "loss": 3.8778,
+      "step": 2134016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5051481855888802e-05,
+      "loss": 3.8708,
+      "step": 2134528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5043095908378283e-05,
+      "loss": 3.8708,
+      "step": 2135040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5034709960867763e-05,
+      "loss": 3.8654,
+      "step": 2135552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5026324013357243e-05,
+      "loss": 3.8733,
+      "step": 2136064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5017970823454186e-05,
+      "loss": 3.874,
+      "step": 2136576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9882256984710693,
+      "eval_runtime": 307.9212,
+      "eval_samples_per_second": 1239.249,
+      "eval_steps_per_second": 38.727,
+      "step": 2136944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5009584875943666e-05,
+      "loss": 3.8755,
+      "step": 2137088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5001198928433146e-05,
+      "loss": 3.8721,
+      "step": 2137600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4992812980922625e-05,
+      "loss": 3.8719,
+      "step": 2138112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4984427033412105e-05,
+      "loss": 3.8821,
+      "step": 2138624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4976041085901585e-05,
+      "loss": 3.883,
+      "step": 2139136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4967655138391064e-05,
+      "loss": 3.8816,
+      "step": 2139648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4959269190880546e-05,
+      "loss": 3.8813,
+      "step": 2140160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4950899622173755e-05,
+      "loss": 3.87,
+      "step": 2140672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4942513674663237e-05,
+      "loss": 3.8657,
+      "step": 2141184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4934144105956446e-05,
+      "loss": 3.8762,
+      "step": 2141696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4925758158445926e-05,
+      "loss": 3.8817,
+      "step": 2142208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4917372210935408e-05,
+      "loss": 3.8702,
+      "step": 2142720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.490900264222862e-05,
+      "loss": 3.8853,
+      "step": 2143232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4900616694718098e-05,
+      "loss": 3.8609,
+      "step": 2143744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4892230747207578e-05,
+      "loss": 3.8704,
+      "step": 2144256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4883844799697058e-05,
+      "loss": 3.8552,
+      "step": 2144768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4875458852186538e-05,
+      "loss": 3.8771,
+      "step": 2145280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4867072904676018e-05,
+      "loss": 3.8778,
+      "step": 2145792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.48586869571655e-05,
+      "loss": 3.8731,
+      "step": 2146304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4850301009654976e-05,
+      "loss": 3.8716,
+      "step": 2146816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4841915062144456e-05,
+      "loss": 3.8876,
+      "step": 2147328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4833529114633938e-05,
+      "loss": 3.8788,
+      "step": 2147840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4825143167123418e-05,
+      "loss": 3.8771,
+      "step": 2148352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4816757219612898e-05,
+      "loss": 3.8731,
+      "step": 2148864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4808387650906109e-05,
+      "loss": 3.884,
+      "step": 2149376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4800001703395589e-05,
+      "loss": 3.8757,
+      "step": 2149888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4791615755885069e-05,
+      "loss": 3.8691,
+      "step": 2150400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4783229808374549e-05,
+      "loss": 3.873,
+      "step": 2150912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.477486023966776e-05,
+      "loss": 3.867,
+      "step": 2151424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.476647429215724e-05,
+      "loss": 3.8667,
+      "step": 2151936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.475810472345045e-05,
+      "loss": 3.8806,
+      "step": 2152448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.474971877593993e-05,
+      "loss": 3.8721,
+      "step": 2152960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.474133282842941e-05,
+      "loss": 3.882,
+      "step": 2153472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4732946880918892e-05,
+      "loss": 3.8866,
+      "step": 2153984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.47245773122121e-05,
+      "loss": 3.8712,
+      "step": 2154496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4716207743505315e-05,
+      "loss": 3.8788,
+      "step": 2155008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4707821795994795e-05,
+      "loss": 3.881,
+      "step": 2155520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4699435848484275e-05,
+      "loss": 3.8637,
+      "step": 2156032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4691049900973753e-05,
+      "loss": 3.8703,
+      "step": 2156544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4682663953463233e-05,
+      "loss": 3.861,
+      "step": 2157056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4674278005952713e-05,
+      "loss": 3.8619,
+      "step": 2157568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4665892058442193e-05,
+      "loss": 3.8746,
+      "step": 2158080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4657506110931673e-05,
+      "loss": 3.885,
+      "step": 2158592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4649120163421155e-05,
+      "loss": 3.8672,
+      "step": 2159104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4640734215910631e-05,
+      "loss": 3.8729,
+      "step": 2159616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4632348268400111e-05,
+      "loss": 3.8716,
+      "step": 2160128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4623962320889593e-05,
+      "loss": 3.8747,
+      "step": 2160640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4615592752182805e-05,
+      "loss": 3.8665,
+      "step": 2161152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4607206804672283e-05,
+      "loss": 3.8538,
+      "step": 2161664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4598820857161763e-05,
+      "loss": 3.8775,
+      "step": 2162176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4590451288454976e-05,
+      "loss": 3.883,
+      "step": 2162688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4582065340944456e-05,
+      "loss": 3.8839,
+      "step": 2163200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4573679393433934e-05,
+      "loss": 3.8661,
+      "step": 2163712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4565293445923414e-05,
+      "loss": 3.8713,
+      "step": 2164224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4556907498412894e-05,
+      "loss": 3.8694,
+      "step": 2164736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4548537929706105e-05,
+      "loss": 3.8597,
+      "step": 2165248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4540151982195585e-05,
+      "loss": 3.885,
+      "step": 2165760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4531766034685065e-05,
+      "loss": 3.8552,
+      "step": 2166272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4523380087174546e-05,
+      "loss": 3.8915,
+      "step": 2166784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4514994139664026e-05,
+      "loss": 3.8727,
+      "step": 2167296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4506608192153506e-05,
+      "loss": 3.86,
+      "step": 2167808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4498238623446717e-05,
+      "loss": 3.8745,
+      "step": 2168320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4489852675936197e-05,
+      "loss": 3.8694,
+      "step": 2168832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4481466728425677e-05,
+      "loss": 3.8553,
+      "step": 2169344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4473080780915157e-05,
+      "loss": 3.8719,
+      "step": 2169856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4464711212208368e-05,
+      "loss": 3.8813,
+      "step": 2170368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4456325264697848e-05,
+      "loss": 3.8652,
+      "step": 2170880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4447955695991059e-05,
+      "loss": 3.8744,
+      "step": 2171392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4439569748480538e-05,
+      "loss": 3.8512,
+      "step": 2171904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4431183800970018e-05,
+      "loss": 3.8673,
+      "step": 2172416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.44227978534595e-05,
+      "loss": 3.8675,
+      "step": 2172928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.441441190594898e-05,
+      "loss": 3.8735,
+      "step": 2173440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.440602595843846e-05,
+      "loss": 3.8673,
+      "step": 2173952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4397640010927938e-05,
+      "loss": 3.868,
+      "step": 2174464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4389254063417418e-05,
+      "loss": 3.8859,
+      "step": 2174976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4380868115906898e-05,
+      "loss": 3.8749,
+      "step": 2175488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4372482168396378e-05,
+      "loss": 3.8717,
+      "step": 2176000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4364096220885858e-05,
+      "loss": 3.868,
+      "step": 2176512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.435571027337534e-05,
+      "loss": 3.8687,
+      "step": 2177024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4347324325864816e-05,
+      "loss": 3.8771,
+      "step": 2177536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4338954757158029e-05,
+      "loss": 3.8656,
+      "step": 2178048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.433058518845124e-05,
+      "loss": 3.871,
+      "step": 2178560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4322215619744454e-05,
+      "loss": 3.8789,
+      "step": 2179072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4313829672233934e-05,
+      "loss": 3.8714,
+      "step": 2179584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.430544372472341e-05,
+      "loss": 3.8522,
+      "step": 2180096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4297057777212892e-05,
+      "loss": 3.8701,
+      "step": 2180608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4288671829702372e-05,
+      "loss": 3.876,
+      "step": 2181120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4280285882191852e-05,
+      "loss": 3.864,
+      "step": 2181632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4271899934681332e-05,
+      "loss": 3.8718,
+      "step": 2182144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4263513987170812e-05,
+      "loss": 3.8696,
+      "step": 2182656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.425512803966029e-05,
+      "loss": 3.8597,
+      "step": 2183168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.424674209214977e-05,
+      "loss": 3.8647,
+      "step": 2183680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.423835614463925e-05,
+      "loss": 3.8669,
+      "step": 2184192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4229970197128732e-05,
+      "loss": 3.877,
+      "step": 2184704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4221584249618212e-05,
+      "loss": 3.8719,
+      "step": 2185216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.421321468091142e-05,
+      "loss": 3.8811,
+      "step": 2185728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4204845112204635e-05,
+      "loss": 3.8674,
+      "step": 2186240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4196459164694115e-05,
+      "loss": 3.8654,
+      "step": 2186752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4188073217183593e-05,
+      "loss": 3.8669,
+      "step": 2187264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4179687269673073e-05,
+      "loss": 3.8783,
+      "step": 2187776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4171301322162553e-05,
+      "loss": 3.874,
+      "step": 2188288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4162931753455764e-05,
+      "loss": 3.8725,
+      "step": 2188800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4154545805945244e-05,
+      "loss": 3.8728,
+      "step": 2189312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4146159858434724e-05,
+      "loss": 3.8662,
+      "step": 2189824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4137790289727936e-05,
+      "loss": 3.8751,
+      "step": 2190336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4129404342217414e-05,
+      "loss": 3.8698,
+      "step": 2190848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4121018394706894e-05,
+      "loss": 3.8658,
+      "step": 2191360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4112632447196374e-05,
+      "loss": 3.87,
+      "step": 2191872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4104246499685856e-05,
+      "loss": 3.8664,
+      "step": 2192384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4095860552175336e-05,
+      "loss": 3.8685,
+      "step": 2192896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4087474604664816e-05,
+      "loss": 3.8651,
+      "step": 2193408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4079088657154296e-05,
+      "loss": 3.8593,
+      "step": 2193920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4070719088447507e-05,
+      "loss": 3.8744,
+      "step": 2194432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4062349519740717e-05,
+      "loss": 3.871,
+      "step": 2194944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4053963572230197e-05,
+      "loss": 3.8703,
+      "step": 2195456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4045577624719677e-05,
+      "loss": 3.8775,
+      "step": 2195968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4037191677209157e-05,
+      "loss": 3.8834,
+      "step": 2196480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4028805729698639e-05,
+      "loss": 3.8692,
+      "step": 2196992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4020419782188119e-05,
+      "loss": 3.867,
+      "step": 2197504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4012033834677595e-05,
+      "loss": 3.8659,
+      "step": 2198016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4003647887167075e-05,
+      "loss": 3.8741,
+      "step": 2198528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3995261939656557e-05,
+      "loss": 3.8684,
+      "step": 2199040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.398689237094977e-05,
+      "loss": 3.8675,
+      "step": 2199552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3978506423439248e-05,
+      "loss": 3.8653,
+      "step": 2200064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3970120475928728e-05,
+      "loss": 3.8772,
+      "step": 2200576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3961734528418208e-05,
+      "loss": 3.8656,
+      "step": 2201088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3953348580907688e-05,
+      "loss": 3.8706,
+      "step": 2201600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3944962633397168e-05,
+      "loss": 3.8716,
+      "step": 2202112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3936593064690378e-05,
+      "loss": 3.8695,
+      "step": 2202624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3928207117179858e-05,
+      "loss": 3.8685,
+      "step": 2203136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.391982116966934e-05,
+      "loss": 3.8798,
+      "step": 2203648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.391143522215882e-05,
+      "loss": 3.8737,
+      "step": 2204160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.39030492746483e-05,
+      "loss": 3.8695,
+      "step": 2204672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3894663327137778e-05,
+      "loss": 3.8762,
+      "step": 2205184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3886277379627258e-05,
+      "loss": 3.881,
+      "step": 2205696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3877891432116738e-05,
+      "loss": 3.8655,
+      "step": 2206208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3869538242213681e-05,
+      "loss": 3.8708,
+      "step": 2206720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3861152294703161e-05,
+      "loss": 3.8723,
+      "step": 2207232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3852766347192641e-05,
+      "loss": 3.8823,
+      "step": 2207744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3844380399682121e-05,
+      "loss": 3.8784,
+      "step": 2208256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.38359944521716e-05,
+      "loss": 3.8769,
+      "step": 2208768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.382760850466108e-05,
+      "loss": 3.8687,
+      "step": 2209280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.381922255715056e-05,
+      "loss": 3.8769,
+      "step": 2209792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3810852988443774e-05,
+      "loss": 3.8751,
+      "step": 2210304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.380246704093325e-05,
+      "loss": 3.8645,
+      "step": 2210816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3794097472226464e-05,
+      "loss": 3.87,
+      "step": 2211328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3785711524715944e-05,
+      "loss": 3.8594,
+      "step": 2211840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3777325577205421e-05,
+      "loss": 3.8734,
+      "step": 2212352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3768939629694903e-05,
+      "loss": 3.8682,
+      "step": 2212864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9870569705963135,
+      "eval_runtime": 308.08,
+      "eval_samples_per_second": 1238.61,
+      "eval_steps_per_second": 38.707,
+      "step": 2213264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3760553682184382e-05,
+      "loss": 3.8789,
+      "step": 2213376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3752167734673862e-05,
+      "loss": 3.8669,
+      "step": 2213888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3743781787163342e-05,
+      "loss": 3.8683,
+      "step": 2214400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3735395839652822e-05,
+      "loss": 3.8811,
+      "step": 2214912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3727009892142304e-05,
+      "loss": 3.8751,
+      "step": 2215424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.371862394463178e-05,
+      "loss": 3.8813,
+      "step": 2215936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.371023799712126e-05,
+      "loss": 3.873,
+      "step": 2216448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3701852049610742e-05,
+      "loss": 3.8685,
+      "step": 2216960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3693466102100222e-05,
+      "loss": 3.8608,
+      "step": 2217472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3685080154589702e-05,
+      "loss": 3.8718,
+      "step": 2217984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3676694207079182e-05,
+      "loss": 3.8723,
+      "step": 2218496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3668308259568664e-05,
+      "loss": 3.8702,
+      "step": 2219008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3659938690861873e-05,
+      "loss": 3.8829,
+      "step": 2219520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3651552743351353e-05,
+      "loss": 3.8545,
+      "step": 2220032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3643166795840834e-05,
+      "loss": 3.8655,
+      "step": 2220544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3634780848330311e-05,
+      "loss": 3.8564,
+      "step": 2221056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3626394900819791e-05,
+      "loss": 3.8735,
+      "step": 2221568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3618008953309273e-05,
+      "loss": 3.8677,
+      "step": 2222080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3609623005798753e-05,
+      "loss": 3.8702,
+      "step": 2222592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3601237058288232e-05,
+      "loss": 3.8748,
+      "step": 2223104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3592851110777712e-05,
+      "loss": 3.8774,
+      "step": 2223616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3584465163267192e-05,
+      "loss": 3.8766,
+      "step": 2224128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.357607921575667e-05,
+      "loss": 3.8702,
+      "step": 2224640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.356769326824615e-05,
+      "loss": 3.8716,
+      "step": 2225152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.355930732073563e-05,
+      "loss": 3.88,
+      "step": 2225664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3550937752028845e-05,
+      "loss": 3.8723,
+      "step": 2226176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3542551804518321e-05,
+      "loss": 3.8612,
+      "step": 2226688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3534165857007803e-05,
+      "loss": 3.8714,
+      "step": 2227200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3525779909497283e-05,
+      "loss": 3.8587,
+      "step": 2227712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3517410340790492e-05,
+      "loss": 3.8665,
+      "step": 2228224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3509040772083706e-05,
+      "loss": 3.8757,
+      "step": 2228736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3500654824573186e-05,
+      "loss": 3.8672,
+      "step": 2229248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3492285255866397e-05,
+      "loss": 3.8753,
+      "step": 2229760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3483899308355877e-05,
+      "loss": 3.8819,
+      "step": 2230272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3475513360845357e-05,
+      "loss": 3.8698,
+      "step": 2230784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3467143792138568e-05,
+      "loss": 3.878,
+      "step": 2231296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3458757844628048e-05,
+      "loss": 3.8786,
+      "step": 2231808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3450371897117528e-05,
+      "loss": 3.8578,
+      "step": 2232320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3441985949607007e-05,
+      "loss": 3.8675,
+      "step": 2232832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3433600002096489e-05,
+      "loss": 3.8547,
+      "step": 2233344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3425214054585966e-05,
+      "loss": 3.8635,
+      "step": 2233856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3416828107075446e-05,
+      "loss": 3.8673,
+      "step": 2234368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3408442159564927e-05,
+      "loss": 3.8826,
+      "step": 2234880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3400056212054407e-05,
+      "loss": 3.8604,
+      "step": 2235392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3391670264543887e-05,
+      "loss": 3.8679,
+      "step": 2235904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3383284317033367e-05,
+      "loss": 3.8674,
+      "step": 2236416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3374898369522849e-05,
+      "loss": 3.87,
+      "step": 2236928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3366512422012325e-05,
+      "loss": 3.8651,
+      "step": 2237440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3358126474501805e-05,
+      "loss": 3.848,
+      "step": 2237952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3349740526991287e-05,
+      "loss": 3.8702,
+      "step": 2238464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3341354579480767e-05,
+      "loss": 3.885,
+      "step": 2238976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3332985010773976e-05,
+      "loss": 3.874,
+      "step": 2239488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3324599063263458e-05,
+      "loss": 3.8619,
+      "step": 2240000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3316213115752938e-05,
+      "loss": 3.865,
+      "step": 2240512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3307827168242418e-05,
+      "loss": 3.8711,
+      "step": 2241024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3299457599535628e-05,
+      "loss": 3.8554,
+      "step": 2241536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3291071652025108e-05,
+      "loss": 3.885,
+      "step": 2242048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3282685704514588e-05,
+      "loss": 3.8518,
+      "step": 2242560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3274299757004068e-05,
+      "loss": 3.8864,
+      "step": 2243072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3265930188297279e-05,
+      "loss": 3.8714,
+      "step": 2243584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3257544240786759e-05,
+      "loss": 3.8552,
+      "step": 2244096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3249158293276239e-05,
+      "loss": 3.8657,
+      "step": 2244608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.324077234576572e-05,
+      "loss": 3.8678,
+      "step": 2245120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.32323863982552e-05,
+      "loss": 3.8516,
+      "step": 2245632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3224000450744677e-05,
+      "loss": 3.8644,
+      "step": 2246144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3215630882037891e-05,
+      "loss": 3.8815,
+      "step": 2246656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3207244934527371e-05,
+      "loss": 3.8624,
+      "step": 2247168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3198858987016851e-05,
+      "loss": 3.8694,
+      "step": 2247680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.319047303950633e-05,
+      "loss": 3.8464,
+      "step": 2248192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.318208709199581e-05,
+      "loss": 3.8629,
+      "step": 2248704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3173717523289022e-05,
+      "loss": 3.8632,
+      "step": 2249216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3165331575778504e-05,
+      "loss": 3.8711,
+      "step": 2249728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.315694562826798e-05,
+      "loss": 3.8639,
+      "step": 2250240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.314855968075746e-05,
+      "loss": 3.867,
+      "step": 2250752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3140173733246942e-05,
+      "loss": 3.8779,
+      "step": 2251264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3131804164540151e-05,
+      "loss": 3.8701,
+      "step": 2251776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.312341821702963e-05,
+      "loss": 3.8669,
+      "step": 2252288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3115032269519112e-05,
+      "loss": 3.8657,
+      "step": 2252800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3106646322008592e-05,
+      "loss": 3.8663,
+      "step": 2253312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3098276753301803e-05,
+      "loss": 3.8718,
+      "step": 2253824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3089890805791283e-05,
+      "loss": 3.8598,
+      "step": 2254336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3081504858280763e-05,
+      "loss": 3.8706,
+      "step": 2254848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3073118910770243e-05,
+      "loss": 3.8787,
+      "step": 2255360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3064732963259723e-05,
+      "loss": 3.8671,
+      "step": 2255872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3056347015749205e-05,
+      "loss": 3.8457,
+      "step": 2256384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3047977447042414e-05,
+      "loss": 3.8687,
+      "step": 2256896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3039591499531895e-05,
+      "loss": 3.871,
+      "step": 2257408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3031205552021375e-05,
+      "loss": 3.8605,
+      "step": 2257920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3022819604510855e-05,
+      "loss": 3.8682,
+      "step": 2258432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3014450035804066e-05,
+      "loss": 3.8662,
+      "step": 2258944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3006064088293546e-05,
+      "loss": 3.8544,
+      "step": 2259456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2997678140783026e-05,
+      "loss": 3.86,
+      "step": 2259968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2989292193272506e-05,
+      "loss": 3.8612,
+      "step": 2260480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2980906245761984e-05,
+      "loss": 3.875,
+      "step": 2260992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2972520298251464e-05,
+      "loss": 3.8688,
+      "step": 2261504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2964150729544677e-05,
+      "loss": 3.8742,
+      "step": 2262016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2955764782034158e-05,
+      "loss": 3.8628,
+      "step": 2262528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2947378834523635e-05,
+      "loss": 3.8622,
+      "step": 2263040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2938992887013115e-05,
+      "loss": 3.8653,
+      "step": 2263552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2930606939502597e-05,
+      "loss": 3.8723,
+      "step": 2264064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2922220991992076e-05,
+      "loss": 3.8698,
+      "step": 2264576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2913835044481556e-05,
+      "loss": 3.8699,
+      "step": 2265088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2905449096971036e-05,
+      "loss": 3.8722,
+      "step": 2265600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2897063149460515e-05,
+      "loss": 3.8607,
+      "step": 2266112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2888693580753727e-05,
+      "loss": 3.8725,
+      "step": 2266624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2880307633243207e-05,
+      "loss": 3.8614,
+      "step": 2267136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2871938064536418e-05,
+      "loss": 3.8641,
+      "step": 2267648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2863552117025898e-05,
+      "loss": 3.8651,
+      "step": 2268160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2855166169515378e-05,
+      "loss": 3.8607,
+      "step": 2268672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.284678022200486e-05,
+      "loss": 3.8606,
+      "step": 2269184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.283839427449434e-05,
+      "loss": 3.8648,
+      "step": 2269696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2830008326983816e-05,
+      "loss": 3.854,
+      "step": 2270208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.282163875827703e-05,
+      "loss": 3.8706,
+      "step": 2270720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.281325281076651e-05,
+      "loss": 3.869,
+      "step": 2271232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2804866863255988e-05,
+      "loss": 3.8626,
+      "step": 2271744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2796480915745468e-05,
+      "loss": 3.8751,
+      "step": 2272256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.278811134703868e-05,
+      "loss": 3.8814,
+      "step": 2272768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.277972539952816e-05,
+      "loss": 3.8647,
+      "step": 2273280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2771355830821372e-05,
+      "loss": 3.8644,
+      "step": 2273792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2762969883310852e-05,
+      "loss": 3.8637,
+      "step": 2274304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2754583935800331e-05,
+      "loss": 3.8673,
+      "step": 2274816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.274619798828981e-05,
+      "loss": 3.8652,
+      "step": 2275328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.273781204077929e-05,
+      "loss": 3.8644,
+      "step": 2275840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.272942609326877e-05,
+      "loss": 3.8618,
+      "step": 2276352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2721040145758251e-05,
+      "loss": 3.8714,
+      "step": 2276864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2712654198247731e-05,
+      "loss": 3.8642,
+      "step": 2277376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2704268250737211e-05,
+      "loss": 3.8677,
+      "step": 2277888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2695882303226691e-05,
+      "loss": 3.8652,
+      "step": 2278400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2687512734519902e-05,
+      "loss": 3.8655,
+      "step": 2278912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2679126787009382e-05,
+      "loss": 3.8616,
+      "step": 2279424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2670740839498862e-05,
+      "loss": 3.8803,
+      "step": 2279936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2662354891988343e-05,
+      "loss": 3.8722,
+      "step": 2280448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2653985323281553e-05,
+      "loss": 3.8671,
+      "step": 2280960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2645599375771033e-05,
+      "loss": 3.8672,
+      "step": 2281472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2637213428260514e-05,
+      "loss": 3.8753,
+      "step": 2281984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.262882748074999e-05,
+      "loss": 3.8669,
+      "step": 2282496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.262044153323947e-05,
+      "loss": 3.8672,
+      "step": 2283008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2612055585728952e-05,
+      "loss": 3.8675,
+      "step": 2283520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2603669638218432e-05,
+      "loss": 3.8808,
+      "step": 2284032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2595283690707912e-05,
+      "loss": 3.874,
+      "step": 2284544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2586930500804856e-05,
+      "loss": 3.872,
+      "step": 2285056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2578544553294336e-05,
+      "loss": 3.8666,
+      "step": 2285568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2570158605783816e-05,
+      "loss": 3.8754,
+      "step": 2286080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2561772658273294e-05,
+      "loss": 3.867,
+      "step": 2286592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2553386710762774e-05,
+      "loss": 3.8589,
+      "step": 2287104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2545000763252254e-05,
+      "loss": 3.8667,
+      "step": 2287616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2536614815741735e-05,
+      "loss": 3.8544,
+      "step": 2288128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2528228868231215e-05,
+      "loss": 3.87,
+      "step": 2288640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2519859299524424e-05,
+      "loss": 3.8628,
+      "step": 2289152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9859445095062256,
+      "eval_runtime": 296.6578,
+      "eval_samples_per_second": 1286.3,
+      "eval_steps_per_second": 40.198,
+      "step": 2289584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2511473352013906e-05,
+      "loss": 3.8811,
+      "step": 2289664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2503087404503386e-05,
+      "loss": 3.8621,
+      "step": 2290176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2494701456992864e-05,
+      "loss": 3.8625,
+      "step": 2290688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2486315509482344e-05,
+      "loss": 3.8803,
+      "step": 2291200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2477929561971826e-05,
+      "loss": 3.8699,
+      "step": 2291712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2469543614461306e-05,
+      "loss": 3.88,
+      "step": 2292224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2461157666950784e-05,
+      "loss": 3.8662,
+      "step": 2292736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2452771719440266e-05,
+      "loss": 3.8633,
+      "step": 2293248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2444385771929746e-05,
+      "loss": 3.8599,
+      "step": 2293760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2435999824419224e-05,
+      "loss": 3.871,
+      "step": 2294272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2427613876908704e-05,
+      "loss": 3.8682,
+      "step": 2294784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2419227929398186e-05,
+      "loss": 3.8661,
+      "step": 2295296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2410858360691396e-05,
+      "loss": 3.8782,
+      "step": 2295808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2402472413180875e-05,
+      "loss": 3.8544,
+      "step": 2296320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2394086465670356e-05,
+      "loss": 3.8611,
+      "step": 2296832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2385700518159836e-05,
+      "loss": 3.855,
+      "step": 2297344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2377314570649314e-05,
+      "loss": 3.8645,
+      "step": 2297856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2368928623138794e-05,
+      "loss": 3.8693,
+      "step": 2298368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2360542675628276e-05,
+      "loss": 3.8632,
+      "step": 2298880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2352156728117754e-05,
+      "loss": 3.8699,
+      "step": 2299392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2343770780607234e-05,
+      "loss": 3.8741,
+      "step": 2299904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2335384833096714e-05,
+      "loss": 3.8695,
+      "step": 2300416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2326998885586196e-05,
+      "loss": 3.871,
+      "step": 2300928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2318612938075674e-05,
+      "loss": 3.8649,
+      "step": 2301440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2310226990565154e-05,
+      "loss": 3.878,
+      "step": 2301952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2301857421858367e-05,
+      "loss": 3.8675,
+      "step": 2302464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2293487853151577e-05,
+      "loss": 3.8617,
+      "step": 2302976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2285101905641057e-05,
+      "loss": 3.8683,
+      "step": 2303488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2276715958130537e-05,
+      "loss": 3.8546,
+      "step": 2304000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2268330010620017e-05,
+      "loss": 3.8627,
+      "step": 2304512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2259944063109497e-05,
+      "loss": 3.8661,
+      "step": 2305024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2251558115598977e-05,
+      "loss": 3.868,
+      "step": 2305536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2243188546892188e-05,
+      "loss": 3.8708,
+      "step": 2306048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2234802599381668e-05,
+      "loss": 3.8771,
+      "step": 2306560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2226433030674879e-05,
+      "loss": 3.8662,
+      "step": 2307072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2218063461968091e-05,
+      "loss": 3.8745,
+      "step": 2307584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2209677514457571e-05,
+      "loss": 3.8719,
+      "step": 2308096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2201291566947051e-05,
+      "loss": 3.8551,
+      "step": 2308608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.219290561943653e-05,
+      "loss": 3.8624,
+      "step": 2309120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2184519671926011e-05,
+      "loss": 3.8524,
+      "step": 2309632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2176133724415491e-05,
+      "loss": 3.8615,
+      "step": 2310144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.216774777690497e-05,
+      "loss": 3.8608,
+      "step": 2310656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.215936182939445e-05,
+      "loss": 3.8783,
+      "step": 2311168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.215097588188393e-05,
+      "loss": 3.8574,
+      "step": 2311680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2142589934373409e-05,
+      "loss": 3.862,
+      "step": 2312192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2134203986862889e-05,
+      "loss": 3.8683,
+      "step": 2312704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.212581803935237e-05,
+      "loss": 3.8633,
+      "step": 2313216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2117448470645581e-05,
+      "loss": 3.8626,
+      "step": 2313728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.210906252313506e-05,
+      "loss": 3.8436,
+      "step": 2314240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2100676575624541e-05,
+      "loss": 3.8685,
+      "step": 2314752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2092307006917752e-05,
+      "loss": 3.8781,
+      "step": 2315264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2083921059407232e-05,
+      "loss": 3.8724,
+      "step": 2315776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2075535111896712e-05,
+      "loss": 3.8617,
+      "step": 2316288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2067149164386192e-05,
+      "loss": 3.8584,
+      "step": 2316800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2058763216875672e-05,
+      "loss": 3.8713,
+      "step": 2317312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2050393648168883e-05,
+      "loss": 3.8472,
+      "step": 2317824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2042007700658363e-05,
+      "loss": 3.879,
+      "step": 2318336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2033621753147843e-05,
+      "loss": 3.8478,
+      "step": 2318848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2025235805637323e-05,
+      "loss": 3.8831,
+      "step": 2319360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2016849858126803e-05,
+      "loss": 3.8679,
+      "step": 2319872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2008463910616283e-05,
+      "loss": 3.8479,
+      "step": 2320384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2000077963105763e-05,
+      "loss": 3.8617,
+      "step": 2320896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1991692015595242e-05,
+      "loss": 3.8676,
+      "step": 2321408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1983306068084722e-05,
+      "loss": 3.8495,
+      "step": 2321920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1974920120574202e-05,
+      "loss": 3.8573,
+      "step": 2322432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1966550551867413e-05,
+      "loss": 3.8755,
+      "step": 2322944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1958164604356893e-05,
+      "loss": 3.8624,
+      "step": 2323456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1949778656846373e-05,
+      "loss": 3.8638,
+      "step": 2323968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1941392709335853e-05,
+      "loss": 3.8455,
+      "step": 2324480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1933006761825333e-05,
+      "loss": 3.8614,
+      "step": 2324992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1924637193118544e-05,
+      "loss": 3.8538,
+      "step": 2325504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1916267624411756e-05,
+      "loss": 3.8713,
+      "step": 2326016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1907881676901236e-05,
+      "loss": 3.8563,
+      "step": 2326528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1899495729390715e-05,
+      "loss": 3.866,
+      "step": 2327040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1891109781880196e-05,
+      "loss": 3.873,
+      "step": 2327552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1882723834369676e-05,
+      "loss": 3.8647,
+      "step": 2328064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1874337886859154e-05,
+      "loss": 3.8625,
+      "step": 2328576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1865968318152367e-05,
+      "loss": 3.8636,
+      "step": 2329088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1857582370641847e-05,
+      "loss": 3.8619,
+      "step": 2329600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1849196423131327e-05,
+      "loss": 3.8676,
+      "step": 2330112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1840810475620807e-05,
+      "loss": 3.8589,
+      "step": 2330624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1832424528110287e-05,
+      "loss": 3.8638,
+      "step": 2331136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1824038580599767e-05,
+      "loss": 3.8699,
+      "step": 2331648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1815652633089245e-05,
+      "loss": 3.8652,
+      "step": 2332160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1807266685578727e-05,
+      "loss": 3.8461,
+      "step": 2332672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1798897116871937e-05,
+      "loss": 3.8645,
+      "step": 2333184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1790511169361417e-05,
+      "loss": 3.8621,
+      "step": 2333696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1782125221850897e-05,
+      "loss": 3.8633,
+      "step": 2334208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1773739274340377e-05,
+      "loss": 3.8602,
+      "step": 2334720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1765353326829857e-05,
+      "loss": 3.8672,
+      "step": 2335232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1756967379319337e-05,
+      "loss": 3.8538,
+      "step": 2335744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1748581431808817e-05,
+      "loss": 3.8568,
+      "step": 2336256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1740195484298297e-05,
+      "loss": 3.8558,
+      "step": 2336768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1731809536787775e-05,
+      "loss": 3.8699,
+      "step": 2337280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1723423589277257e-05,
+      "loss": 3.864,
+      "step": 2337792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1715054020570468e-05,
+      "loss": 3.8734,
+      "step": 2338304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.170668445186368e-05,
+      "loss": 3.8583,
+      "step": 2338816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1698298504353158e-05,
+      "loss": 3.8578,
+      "step": 2339328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1689912556842638e-05,
+      "loss": 3.859,
+      "step": 2339840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1681526609332118e-05,
+      "loss": 3.8715,
+      "step": 2340352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.16731406618216e-05,
+      "loss": 3.8674,
+      "step": 2340864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1664754714311078e-05,
+      "loss": 3.8667,
+      "step": 2341376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1656368766800558e-05,
+      "loss": 3.8637,
+      "step": 2341888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1647982819290038e-05,
+      "loss": 3.8578,
+      "step": 2342400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1639596871779518e-05,
+      "loss": 3.8709,
+      "step": 2342912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1631210924268998e-05,
+      "loss": 3.8578,
+      "step": 2343424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1622824976758478e-05,
+      "loss": 3.8575,
+      "step": 2343936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.161445540805169e-05,
+      "loss": 3.8633,
+      "step": 2344448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1606069460541169e-05,
+      "loss": 3.8551,
+      "step": 2344960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1597699891834381e-05,
+      "loss": 3.861,
+      "step": 2345472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1589313944323861e-05,
+      "loss": 3.8642,
+      "step": 2345984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.158092799681334e-05,
+      "loss": 3.8499,
+      "step": 2346496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.157254204930282e-05,
+      "loss": 3.8672,
+      "step": 2347008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1564156101792301e-05,
+      "loss": 3.8633,
+      "step": 2347520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.155577015428178e-05,
+      "loss": 3.8592,
+      "step": 2348032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.154738420677126e-05,
+      "loss": 3.8739,
+      "step": 2348544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.153899825926074e-05,
+      "loss": 3.8771,
+      "step": 2349056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1530612311750221e-05,
+      "loss": 3.8635,
+      "step": 2349568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.15222263642397e-05,
+      "loss": 3.8597,
+      "step": 2350080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1513856795532912e-05,
+      "loss": 3.8564,
+      "step": 2350592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1505470848022392e-05,
+      "loss": 3.8644,
+      "step": 2351104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.149708490051187e-05,
+      "loss": 3.8576,
+      "step": 2351616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1488715331805082e-05,
+      "loss": 3.864,
+      "step": 2352128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1480329384294562e-05,
+      "loss": 3.8568,
+      "step": 2352640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1471943436784042e-05,
+      "loss": 3.8705,
+      "step": 2353152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1463557489273522e-05,
+      "loss": 3.8616,
+      "step": 2353664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1455171541763002e-05,
+      "loss": 3.8611,
+      "step": 2354176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1446785594252482e-05,
+      "loss": 3.8632,
+      "step": 2354688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.143839964674196e-05,
+      "loss": 3.8628,
+      "step": 2355200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1430013699231442e-05,
+      "loss": 3.856,
+      "step": 2355712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1421627751720922e-05,
+      "loss": 3.8743,
+      "step": 2356224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1413241804210402e-05,
+      "loss": 3.8716,
+      "step": 2356736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.140485585669988e-05,
+      "loss": 3.8677,
+      "step": 2357248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1396469909189362e-05,
+      "loss": 3.8623,
+      "step": 2357760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1388116719286304e-05,
+      "loss": 3.8663,
+      "step": 2358272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1379730771775785e-05,
+      "loss": 3.8618,
+      "step": 2358784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1371344824265263e-05,
+      "loss": 3.8644,
+      "step": 2359296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1362958876754743e-05,
+      "loss": 3.864,
+      "step": 2359808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1354572929244223e-05,
+      "loss": 3.8806,
+      "step": 2360320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1346186981733703e-05,
+      "loss": 3.8657,
+      "step": 2360832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1337801034223183e-05,
+      "loss": 3.8728,
+      "step": 2361344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1329415086712663e-05,
+      "loss": 3.8619,
+      "step": 2361856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1321029139202143e-05,
+      "loss": 3.8713,
+      "step": 2362368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1312659570495354e-05,
+      "loss": 3.8625,
+      "step": 2362880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1304273622984834e-05,
+      "loss": 3.8568,
+      "step": 2363392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1295887675474316e-05,
+      "loss": 3.8651,
+      "step": 2363904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1287518106767525e-05,
+      "loss": 3.8495,
+      "step": 2364416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1279132159257005e-05,
+      "loss": 3.8656,
+      "step": 2364928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1270762590550217e-05,
+      "loss": 3.8627,
+      "step": 2365440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.984827995300293,
+      "eval_runtime": 320.7907,
+      "eval_samples_per_second": 1189.533,
+      "eval_steps_per_second": 37.174,
+      "step": 2365904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1262376643039697e-05,
+      "loss": 3.8611,
+      "step": 2365952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1253990695529177e-05,
+      "loss": 3.8652,
+      "step": 2366464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1245604748018657e-05,
+      "loss": 3.8593,
+      "step": 2366976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1237218800508137e-05,
+      "loss": 3.8707,
+      "step": 2367488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1228832852997615e-05,
+      "loss": 3.867,
+      "step": 2368000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1220446905487097e-05,
+      "loss": 3.8757,
+      "step": 2368512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1212060957976577e-05,
+      "loss": 3.8637,
+      "step": 2369024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1203675010466057e-05,
+      "loss": 3.8594,
+      "step": 2369536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1195289062955535e-05,
+      "loss": 3.8525,
+      "step": 2370048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1186903115445017e-05,
+      "loss": 3.8674,
+      "step": 2370560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1178517167934497e-05,
+      "loss": 3.8637,
+      "step": 2371072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1170131220423975e-05,
+      "loss": 3.8651,
+      "step": 2371584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1161745272913455e-05,
+      "loss": 3.8683,
+      "step": 2372096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1153375704206667e-05,
+      "loss": 3.8553,
+      "step": 2372608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1144989756696147e-05,
+      "loss": 3.8545,
+      "step": 2373120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1136620187989358e-05,
+      "loss": 3.8521,
+      "step": 2373632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1128234240478838e-05,
+      "loss": 3.8575,
+      "step": 2374144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1119848292968318e-05,
+      "loss": 3.8663,
+      "step": 2374656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1111462345457798e-05,
+      "loss": 3.8626,
+      "step": 2375168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1103076397947278e-05,
+      "loss": 3.8608,
+      "step": 2375680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1094690450436758e-05,
+      "loss": 3.874,
+      "step": 2376192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1086304502926238e-05,
+      "loss": 3.8655,
+      "step": 2376704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1077918555415718e-05,
+      "loss": 3.8676,
+      "step": 2377216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1069532607905198e-05,
+      "loss": 3.8641,
+      "step": 2377728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1061146660394678e-05,
+      "loss": 3.8753,
+      "step": 2378240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1052760712884156e-05,
+      "loss": 3.8584,
+      "step": 2378752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1044374765373638e-05,
+      "loss": 3.862,
+      "step": 2379264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1035988817863118e-05,
+      "loss": 3.8587,
+      "step": 2379776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1027602870352596e-05,
+      "loss": 3.8593,
+      "step": 2380288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1019216922842077e-05,
+      "loss": 3.8575,
+      "step": 2380800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1010847354135288e-05,
+      "loss": 3.861,
+      "step": 2381312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.10024777854285e-05,
+      "loss": 3.8654,
+      "step": 2381824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.099410821672171e-05,
+      "loss": 3.8683,
+      "step": 2382336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.098572226921119e-05,
+      "loss": 3.8704,
+      "step": 2382848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0977336321700671e-05,
+      "loss": 3.8627,
+      "step": 2383360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0968966752993882e-05,
+      "loss": 3.8707,
+      "step": 2383872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0960580805483362e-05,
+      "loss": 3.8669,
+      "step": 2384384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0952194857972842e-05,
+      "loss": 3.8524,
+      "step": 2384896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0943808910462322e-05,
+      "loss": 3.8595,
+      "step": 2385408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.09354229629518e-05,
+      "loss": 3.8472,
+      "step": 2385920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0927037015441282e-05,
+      "loss": 3.8584,
+      "step": 2386432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0918651067930762e-05,
+      "loss": 3.8565,
+      "step": 2386944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0910265120420242e-05,
+      "loss": 3.8749,
+      "step": 2387456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.090187917290972e-05,
+      "loss": 3.8546,
+      "step": 2387968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0893509604202933e-05,
+      "loss": 3.8556,
+      "step": 2388480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0885123656692413e-05,
+      "loss": 3.8605,
+      "step": 2388992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0876737709181891e-05,
+      "loss": 3.8649,
+      "step": 2389504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0868351761671373e-05,
+      "loss": 3.8605,
+      "step": 2390016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0859965814160852e-05,
+      "loss": 3.8423,
+      "step": 2390528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0851579866650332e-05,
+      "loss": 3.8607,
+      "step": 2391040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0843193919139812e-05,
+      "loss": 3.8753,
+      "step": 2391552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0834824350433023e-05,
+      "loss": 3.8646,
+      "step": 2392064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0826438402922503e-05,
+      "loss": 3.8593,
+      "step": 2392576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0818052455411983e-05,
+      "loss": 3.8539,
+      "step": 2393088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0809666507901463e-05,
+      "loss": 3.8672,
+      "step": 2393600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0801280560390943e-05,
+      "loss": 3.8476,
+      "step": 2394112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0792894612880423e-05,
+      "loss": 3.8737,
+      "step": 2394624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0784508665369903e-05,
+      "loss": 3.8433,
+      "step": 2395136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0776122717859383e-05,
+      "loss": 3.8795,
+      "step": 2395648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0767736770348863e-05,
+      "loss": 3.8657,
+      "step": 2396160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0759350822838341e-05,
+      "loss": 3.8465,
+      "step": 2396672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0750981254131554e-05,
+      "loss": 3.8567,
+      "step": 2397184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0742595306621034e-05,
+      "loss": 3.8701,
+      "step": 2397696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0734209359110513e-05,
+      "loss": 3.8416,
+      "step": 2398208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0725823411599993e-05,
+      "loss": 3.851,
+      "step": 2398720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0717437464089473e-05,
+      "loss": 3.8746,
+      "step": 2399232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0709051516578953e-05,
+      "loss": 3.8606,
+      "step": 2399744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0700665569068433e-05,
+      "loss": 3.8585,
+      "step": 2400256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0692279621557913e-05,
+      "loss": 3.8426,
+      "step": 2400768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0683893674047393e-05,
+      "loss": 3.8553,
+      "step": 2401280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0675524105340606e-05,
+      "loss": 3.8512,
+      "step": 2401792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0667138157830084e-05,
+      "loss": 3.8695,
+      "step": 2402304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0658768589123295e-05,
+      "loss": 3.8526,
+      "step": 2402816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0650382641612776e-05,
+      "loss": 3.8632,
+      "step": 2403328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0641996694102255e-05,
+      "loss": 3.8661,
+      "step": 2403840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0633610746591735e-05,
+      "loss": 3.8616,
+      "step": 2404352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0625224799081215e-05,
+      "loss": 3.8608,
+      "step": 2404864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0616838851570696e-05,
+      "loss": 3.8597,
+      "step": 2405376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0608452904060174e-05,
+      "loss": 3.863,
+      "step": 2405888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0600066956549654e-05,
+      "loss": 3.863,
+      "step": 2406400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0591713766646598e-05,
+      "loss": 3.8546,
+      "step": 2406912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0583327819136078e-05,
+      "loss": 3.8613,
+      "step": 2407424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0574941871625558e-05,
+      "loss": 3.8642,
+      "step": 2407936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0566555924115038e-05,
+      "loss": 3.8631,
+      "step": 2408448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0558169976604518e-05,
+      "loss": 3.8432,
+      "step": 2408960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0549784029093998e-05,
+      "loss": 3.8589,
+      "step": 2409472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0541414460387208e-05,
+      "loss": 3.858,
+      "step": 2409984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0533028512876688e-05,
+      "loss": 3.8602,
+      "step": 2410496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0524642565366168e-05,
+      "loss": 3.8596,
+      "step": 2411008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0516256617855648e-05,
+      "loss": 3.8599,
+      "step": 2411520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0507870670345128e-05,
+      "loss": 3.8525,
+      "step": 2412032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0499484722834608e-05,
+      "loss": 3.8501,
+      "step": 2412544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0491115154127819e-05,
+      "loss": 3.8535,
+      "step": 2413056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0482729206617299e-05,
+      "loss": 3.8685,
+      "step": 2413568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0474343259106779e-05,
+      "loss": 3.857,
+      "step": 2414080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.046595731159626e-05,
+      "loss": 3.8692,
+      "step": 2414592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0457571364085739e-05,
+      "loss": 3.8577,
+      "step": 2415104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.044920179537895e-05,
+      "loss": 3.8564,
+      "step": 2415616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0440815847868431e-05,
+      "loss": 3.8545,
+      "step": 2416128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.043242990035791e-05,
+      "loss": 3.8686,
+      "step": 2416640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.042404395284739e-05,
+      "loss": 3.861,
+      "step": 2417152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0415674384140602e-05,
+      "loss": 3.8642,
+      "step": 2417664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0407288436630082e-05,
+      "loss": 3.8605,
+      "step": 2418176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.039890248911956e-05,
+      "loss": 3.852,
+      "step": 2418688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0390516541609042e-05,
+      "loss": 3.8725,
+      "step": 2419200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0382130594098522e-05,
+      "loss": 3.8485,
+      "step": 2419712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0373744646588e-05,
+      "loss": 3.8561,
+      "step": 2420224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.036535869907748e-05,
+      "loss": 3.8635,
+      "step": 2420736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0356989130370692e-05,
+      "loss": 3.8556,
+      "step": 2421248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0348603182860172e-05,
+      "loss": 3.8528,
+      "step": 2421760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0340217235349652e-05,
+      "loss": 3.8596,
+      "step": 2422272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0331847666642863e-05,
+      "loss": 3.8479,
+      "step": 2422784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0323461719132343e-05,
+      "loss": 3.8634,
+      "step": 2423296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0315075771621823e-05,
+      "loss": 3.857,
+      "step": 2423808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0306689824111303e-05,
+      "loss": 3.8594,
+      "step": 2424320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0298303876600783e-05,
+      "loss": 3.8662,
+      "step": 2424832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0289917929090263e-05,
+      "loss": 3.8761,
+      "step": 2425344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0281531981579743e-05,
+      "loss": 3.8619,
+      "step": 2425856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0273146034069223e-05,
+      "loss": 3.855,
+      "step": 2426368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0264760086558703e-05,
+      "loss": 3.8555,
+      "step": 2426880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0256374139048181e-05,
+      "loss": 3.8605,
+      "step": 2427392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0247988191537663e-05,
+      "loss": 3.8587,
+      "step": 2427904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0239618622830873e-05,
+      "loss": 3.8588,
+      "step": 2428416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0231232675320353e-05,
+      "loss": 3.854,
+      "step": 2428928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0222846727809833e-05,
+      "loss": 3.8648,
+      "step": 2429440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0214460780299313e-05,
+      "loss": 3.8571,
+      "step": 2429952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0206074832788793e-05,
+      "loss": 3.863,
+      "step": 2430464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0197688885278273e-05,
+      "loss": 3.8559,
+      "step": 2430976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0189302937767753e-05,
+      "loss": 3.8571,
+      "step": 2431488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0180916990257233e-05,
+      "loss": 3.8543,
+      "step": 2432000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0172547421550446e-05,
+      "loss": 3.8695,
+      "step": 2432512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0164161474039924e-05,
+      "loss": 3.8702,
+      "step": 2433024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0155791905333135e-05,
+      "loss": 3.8666,
+      "step": 2433536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0147405957822616e-05,
+      "loss": 3.8558,
+      "step": 2434048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0139020010312095e-05,
+      "loss": 3.8584,
+      "step": 2434560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0130634062801575e-05,
+      "loss": 3.8628,
+      "step": 2435072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0122248115291054e-05,
+      "loss": 3.86,
+      "step": 2435584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0113862167780536e-05,
+      "loss": 3.8636,
+      "step": 2436096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0105476220270014e-05,
+      "loss": 3.8707,
+      "step": 2436608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0097090272759494e-05,
+      "loss": 3.8624,
+      "step": 2437120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0088737082856438e-05,
+      "loss": 3.8708,
+      "step": 2437632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.008036751414965e-05,
+      "loss": 3.8586,
+      "step": 2438144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0071981566639128e-05,
+      "loss": 3.868,
+      "step": 2438656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0063595619128608e-05,
+      "loss": 3.8602,
+      "step": 2439168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0055209671618088e-05,
+      "loss": 3.8531,
+      "step": 2439680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0046823724107568e-05,
+      "loss": 3.8631,
+      "step": 2440192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0038437776597048e-05,
+      "loss": 3.8463,
+      "step": 2440704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0030051829086528e-05,
+      "loss": 3.8595,
+      "step": 2441216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.002168226037974e-05,
+      "loss": 3.86,
+      "step": 2441728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9838333129882812,
+      "eval_runtime": 295.1585,
+      "eval_samples_per_second": 1292.834,
+      "eval_steps_per_second": 40.402,
+      "step": 2442224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0013296312869219e-05,
+      "loss": 3.8572,
+      "step": 2442240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0004910365358699e-05,
+      "loss": 3.8596,
+      "step": 2442752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.99652441784818e-06,
+      "loss": 3.8561,
+      "step": 2443264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.988138470337659e-06,
+      "loss": 3.8671,
+      "step": 2443776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.979752522827139e-06,
+      "loss": 3.8659,
+      "step": 2444288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.971366575316619e-06,
+      "loss": 3.8736,
+      "step": 2444800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.9629806278061e-06,
+      "loss": 3.8591,
+      "step": 2445312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.95461105909931e-06,
+      "loss": 3.8553,
+      "step": 2445824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.946225111588791e-06,
+      "loss": 3.8522,
+      "step": 2446336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.937839164078271e-06,
+      "loss": 3.8664,
+      "step": 2446848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.92945321656775e-06,
+      "loss": 3.8612,
+      "step": 2447360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.92106726905723e-06,
+      "loss": 3.863,
+      "step": 2447872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.912697700350442e-06,
+      "loss": 3.8644,
+      "step": 2448384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.904328131643653e-06,
+      "loss": 3.8524,
+      "step": 2448896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.895942184133133e-06,
+      "loss": 3.8552,
+      "step": 2449408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.887556236622612e-06,
+      "loss": 3.8523,
+      "step": 2449920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.879170289112092e-06,
+      "loss": 3.8491,
+      "step": 2450432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.870784341601572e-06,
+      "loss": 3.8621,
+      "step": 2450944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.862398394091052e-06,
+      "loss": 3.8589,
+      "step": 2451456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.854012446580532e-06,
+      "loss": 3.8604,
+      "step": 2451968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.845626499070012e-06,
+      "loss": 3.8744,
+      "step": 2452480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.837240551559492e-06,
+      "loss": 3.8571,
+      "step": 2452992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.828870982852703e-06,
+      "loss": 3.8643,
+      "step": 2453504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.820485035342183e-06,
+      "loss": 3.8587,
+      "step": 2454016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.812099087831663e-06,
+      "loss": 3.8709,
+      "step": 2454528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.803713140321143e-06,
+      "loss": 3.857,
+      "step": 2455040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.795327192810623e-06,
+      "loss": 3.8597,
+      "step": 2455552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.786957624103835e-06,
+      "loss": 3.853,
+      "step": 2456064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.778571676593314e-06,
+      "loss": 3.859,
+      "step": 2456576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.770185729082794e-06,
+      "loss": 3.8534,
+      "step": 2457088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.761799781572273e-06,
+      "loss": 3.8595,
+      "step": 2457600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.753413834061753e-06,
+      "loss": 3.8651,
+      "step": 2458112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.745027886551233e-06,
+      "loss": 3.8636,
+      "step": 2458624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.736658317844446e-06,
+      "loss": 3.8653,
+      "step": 2459136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.728272370333926e-06,
+      "loss": 3.862,
+      "step": 2459648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.719902801627135e-06,
+      "loss": 3.8619,
+      "step": 2460160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.711516854116617e-06,
+      "loss": 3.8665,
+      "step": 2460672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.703130906606097e-06,
+      "loss": 3.8493,
+      "step": 2461184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.694744959095576e-06,
+      "loss": 3.8598,
+      "step": 2461696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.686359011585055e-06,
+      "loss": 3.8437,
+      "step": 2462208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.677973064074536e-06,
+      "loss": 3.8575,
+      "step": 2462720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.669587116564016e-06,
+      "loss": 3.8508,
+      "step": 2463232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.661201169053495e-06,
+      "loss": 3.8693,
+      "step": 2463744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.652815221542975e-06,
+      "loss": 3.8577,
+      "step": 2464256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.644429274032456e-06,
+      "loss": 3.8517,
+      "step": 2464768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.636043326521934e-06,
+      "loss": 3.8583,
+      "step": 2465280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.627673757815147e-06,
+      "loss": 3.8628,
+      "step": 2465792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.619287810304627e-06,
+      "loss": 3.8542,
+      "step": 2466304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.610901862794107e-06,
+      "loss": 3.8425,
+      "step": 2466816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.602515915283585e-06,
+      "loss": 3.8533,
+      "step": 2467328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.594129967773067e-06,
+      "loss": 3.8715,
+      "step": 2467840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.585760399066278e-06,
+      "loss": 3.8672,
+      "step": 2468352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.577374451555758e-06,
+      "loss": 3.855,
+      "step": 2468864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.568988504045237e-06,
+      "loss": 3.8552,
+      "step": 2469376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.560602556534717e-06,
+      "loss": 3.8646,
+      "step": 2469888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.552216609024197e-06,
+      "loss": 3.8421,
+      "step": 2470400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.543830661513677e-06,
+      "loss": 3.8692,
+      "step": 2470912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.535461092806888e-06,
+      "loss": 3.8416,
+      "step": 2471424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.527075145296368e-06,
+      "loss": 3.8689,
+      "step": 2471936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.51870557658958e-06,
+      "loss": 3.8663,
+      "step": 2472448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.510319629079059e-06,
+      "loss": 3.8446,
+      "step": 2472960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.501933681568539e-06,
+      "loss": 3.851,
+      "step": 2473472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.49354773405802e-06,
+      "loss": 3.8668,
+      "step": 2473984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.485161786547499e-06,
+      "loss": 3.8359,
+      "step": 2474496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.476775839036979e-06,
+      "loss": 3.8499,
+      "step": 2475008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.468389891526459e-06,
+      "loss": 3.8694,
+      "step": 2475520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.46000394401594e-06,
+      "loss": 3.8598,
+      "step": 2476032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.451617996505419e-06,
+      "loss": 3.8567,
+      "step": 2476544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.443232048994898e-06,
+      "loss": 3.8417,
+      "step": 2477056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.434846101484378e-06,
+      "loss": 3.8519,
+      "step": 2477568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.42647653277759e-06,
+      "loss": 3.8456,
+      "step": 2478080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.41809058526707e-06,
+      "loss": 3.8682,
+      "step": 2478592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.409704637756551e-06,
+      "loss": 3.8471,
+      "step": 2479104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.401335069049762e-06,
+      "loss": 3.8608,
+      "step": 2479616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.39294912153924e-06,
+      "loss": 3.8597,
+      "step": 2480128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.384563174028722e-06,
+      "loss": 3.8638,
+      "step": 2480640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.376177226518202e-06,
+      "loss": 3.8526,
+      "step": 2481152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.36779127900768e-06,
+      "loss": 3.8557,
+      "step": 2481664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.35940533149716e-06,
+      "loss": 3.8634,
+      "step": 2482176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.351019383986641e-06,
+      "loss": 3.8573,
+      "step": 2482688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.342633436476121e-06,
+      "loss": 3.8543,
+      "step": 2483200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.3342474889656e-06,
+      "loss": 3.8534,
+      "step": 2483712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.325861541455081e-06,
+      "loss": 3.862,
+      "step": 2484224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.317491972748292e-06,
+      "loss": 3.8595,
+      "step": 2484736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.30910602523777e-06,
+      "loss": 3.8453,
+      "step": 2485248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.300736456530983e-06,
+      "loss": 3.8532,
+      "step": 2485760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.292350509020463e-06,
+      "loss": 3.8541,
+      "step": 2486272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.283964561509943e-06,
+      "loss": 3.8617,
+      "step": 2486784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.275578613999423e-06,
+      "loss": 3.8544,
+      "step": 2487296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.267192666488903e-06,
+      "loss": 3.8584,
+      "step": 2487808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.258806718978383e-06,
+      "loss": 3.8501,
+      "step": 2488320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.250420771467863e-06,
+      "loss": 3.8454,
+      "step": 2488832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.242034823957342e-06,
+      "loss": 3.8504,
+      "step": 2489344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.233665255250553e-06,
+      "loss": 3.8636,
+      "step": 2489856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.225279307740033e-06,
+      "loss": 3.8563,
+      "step": 2490368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.216893360229513e-06,
+      "loss": 3.8664,
+      "step": 2490880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.208507412718993e-06,
+      "loss": 3.8579,
+      "step": 2491392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.200121465208473e-06,
+      "loss": 3.8498,
+      "step": 2491904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.191751896501684e-06,
+      "loss": 3.8502,
+      "step": 2492416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.183365948991164e-06,
+      "loss": 3.8651,
+      "step": 2492928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.174980001480644e-06,
+      "loss": 3.8572,
+      "step": 2493440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.166594053970125e-06,
+      "loss": 3.8645,
+      "step": 2493952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.158208106459604e-06,
+      "loss": 3.8543,
+      "step": 2494464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.149822158949084e-06,
+      "loss": 3.8506,
+      "step": 2494976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.141436211438564e-06,
+      "loss": 3.8695,
+      "step": 2495488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.133050263928044e-06,
+      "loss": 3.8444,
+      "step": 2496000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.124680695221254e-06,
+      "loss": 3.8539,
+      "step": 2496512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.116294747710736e-06,
+      "loss": 3.858,
+      "step": 2497024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.107908800200216e-06,
+      "loss": 3.8543,
+      "step": 2497536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.099539231493425e-06,
+      "loss": 3.8483,
+      "step": 2498048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.091153283982907e-06,
+      "loss": 3.8533,
+      "step": 2498560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.082767336472387e-06,
+      "loss": 3.8502,
+      "step": 2499072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.074381388961865e-06,
+      "loss": 3.8575,
+      "step": 2499584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.065995441451345e-06,
+      "loss": 3.855,
+      "step": 2500096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.057625872744557e-06,
+      "loss": 3.8534,
+      "step": 2500608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.049239925234037e-06,
+      "loss": 3.864,
+      "step": 2501120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.040853977723517e-06,
+      "loss": 3.8739,
+      "step": 2501632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.032468030212997e-06,
+      "loss": 3.8597,
+      "step": 2502144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.024082082702477e-06,
+      "loss": 3.8519,
+      "step": 2502656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.015696135191955e-06,
+      "loss": 3.8508,
+      "step": 2503168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.007326566485168e-06,
+      "loss": 3.8523,
+      "step": 2503680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.998940618974648e-06,
+      "loss": 3.855,
+      "step": 2504192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.990554671464128e-06,
+      "loss": 3.8562,
+      "step": 2504704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.982168723953608e-06,
+      "loss": 3.8528,
+      "step": 2505216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.973782776443088e-06,
+      "loss": 3.859,
+      "step": 2505728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.965396828932568e-06,
+      "loss": 3.8535,
+      "step": 2506240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.95702726022578e-06,
+      "loss": 3.8565,
+      "step": 2506752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.948641312715258e-06,
+      "loss": 3.8532,
+      "step": 2507264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.940255365204738e-06,
+      "loss": 3.8528,
+      "step": 2507776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.931869417694218e-06,
+      "loss": 3.8514,
+      "step": 2508288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.923483470183698e-06,
+      "loss": 3.867,
+      "step": 2508800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.915113901476909e-06,
+      "loss": 3.8669,
+      "step": 2509312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.906744332770122e-06,
+      "loss": 3.8667,
+      "step": 2509824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.898358385259602e-06,
+      "loss": 3.8479,
+      "step": 2510336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.88997243774908e-06,
+      "loss": 3.8601,
+      "step": 2510848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.881586490238561e-06,
+      "loss": 3.8565,
+      "step": 2511360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.873200542728041e-06,
+      "loss": 3.8594,
+      "step": 2511872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.86481459521752e-06,
+      "loss": 3.8551,
+      "step": 2512384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.856428647707001e-06,
+      "loss": 3.8703,
+      "step": 2512896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.848042700196481e-06,
+      "loss": 3.8542,
+      "step": 2513408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.839656752685961e-06,
+      "loss": 3.8739,
+      "step": 2513920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.83127080517544e-06,
+      "loss": 3.8529,
+      "step": 2514432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.822901236468652e-06,
+      "loss": 3.8657,
+      "step": 2514944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.814515288958132e-06,
+      "loss": 3.8558,
+      "step": 2515456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.80612934144761e-06,
+      "loss": 3.8503,
+      "step": 2515968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.797743393937092e-06,
+      "loss": 3.8607,
+      "step": 2516480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.789357446426572e-06,
+      "loss": 3.8462,
+      "step": 2516992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.780971498916052e-06,
+      "loss": 3.8549,
+      "step": 2517504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.772601930209263e-06,
+      "loss": 3.8591,
+      "step": 2518016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.764215982698743e-06,
+      "loss": 3.8535,
+      "step": 2518528
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.982947826385498,
+      "eval_runtime": 291.8278,
+      "eval_samples_per_second": 1307.59,
+      "eval_steps_per_second": 40.863,
+      "step": 2518544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.755830035188222e-06,
+      "loss": 3.8568,
+      "step": 2519040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.747444087677702e-06,
+      "loss": 3.855,
+      "step": 2519552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.739058140167182e-06,
+      "loss": 3.8622,
+      "step": 2520064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.730672192656662e-06,
+      "loss": 3.8583,
+      "step": 2520576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.722286245146142e-06,
+      "loss": 3.8748,
+      "step": 2521088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.713900297635622e-06,
+      "loss": 3.8537,
+      "step": 2521600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.705514350125102e-06,
+      "loss": 3.8533,
+      "step": 2522112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.697128402614582e-06,
+      "loss": 3.849,
+      "step": 2522624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.68874245510406e-06,
+      "loss": 3.8601,
+      "step": 2523136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.680356507593542e-06,
+      "loss": 3.8565,
+      "step": 2523648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.671970560083022e-06,
+      "loss": 3.8582,
+      "step": 2524160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.6635846125725e-06,
+      "loss": 3.86,
+      "step": 2524672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.655215043865713e-06,
+      "loss": 3.8477,
+      "step": 2525184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.646829096355193e-06,
+      "loss": 3.8584,
+      "step": 2525696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.638443148844673e-06,
+      "loss": 3.8466,
+      "step": 2526208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.630057201334153e-06,
+      "loss": 3.8505,
+      "step": 2526720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.621671253823633e-06,
+      "loss": 3.8563,
+      "step": 2527232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.613285306313113e-06,
+      "loss": 3.8534,
+      "step": 2527744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.60489935880259e-06,
+      "loss": 3.8553,
+      "step": 2528256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.596529790095803e-06,
+      "loss": 3.8735,
+      "step": 2528768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.588143842585283e-06,
+      "loss": 3.8545,
+      "step": 2529280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.579757895074763e-06,
+      "loss": 3.8607,
+      "step": 2529792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.571371947564243e-06,
+      "loss": 3.8554,
+      "step": 2530304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.562986000053723e-06,
+      "loss": 3.864,
+      "step": 2530816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.554600052543203e-06,
+      "loss": 3.8558,
+      "step": 2531328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.546214105032681e-06,
+      "loss": 3.8596,
+      "step": 2531840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.537828157522163e-06,
+      "loss": 3.8503,
+      "step": 2532352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.529442210011643e-06,
+      "loss": 3.8579,
+      "step": 2532864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.521056262501121e-06,
+      "loss": 3.8484,
+      "step": 2533376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.512670314990603e-06,
+      "loss": 3.8572,
+      "step": 2533888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.504300746283814e-06,
+      "loss": 3.8598,
+      "step": 2534400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.495914798773294e-06,
+      "loss": 3.8576,
+      "step": 2534912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.487528851262774e-06,
+      "loss": 3.8659,
+      "step": 2535424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.479159282555984e-06,
+      "loss": 3.8567,
+      "step": 2535936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.470789713849197e-06,
+      "loss": 3.8606,
+      "step": 2536448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.462403766338677e-06,
+      "loss": 3.8612,
+      "step": 2536960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.454017818828155e-06,
+      "loss": 3.8516,
+      "step": 2537472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.445631871317635e-06,
+      "loss": 3.8528,
+      "step": 2537984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.437245923807117e-06,
+      "loss": 3.8401,
+      "step": 2538496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.428859976296595e-06,
+      "loss": 3.8538,
+      "step": 2539008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.420474028786075e-06,
+      "loss": 3.8516,
+      "step": 2539520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.412088081275555e-06,
+      "loss": 3.8653,
+      "step": 2540032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.403702133765036e-06,
+      "loss": 3.8523,
+      "step": 2540544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.395316186254515e-06,
+      "loss": 3.8517,
+      "step": 2541056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.386930238743995e-06,
+      "loss": 3.8574,
+      "step": 2541568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.378544291233476e-06,
+      "loss": 3.8572,
+      "step": 2542080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.370158343722955e-06,
+      "loss": 3.8535,
+      "step": 2542592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.361772396212435e-06,
+      "loss": 3.8407,
+      "step": 2543104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.353386448701915e-06,
+      "loss": 3.8471,
+      "step": 2543616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.345000501191394e-06,
+      "loss": 3.8694,
+      "step": 2544128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.336630932484605e-06,
+      "loss": 3.8657,
+      "step": 2544640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.328244984974085e-06,
+      "loss": 3.8493,
+      "step": 2545152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.319875416267298e-06,
+      "loss": 3.851,
+      "step": 2545664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.311489468756776e-06,
+      "loss": 3.8652,
+      "step": 2546176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.303119900049988e-06,
+      "loss": 3.8407,
+      "step": 2546688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.294733952539468e-06,
+      "loss": 3.8618,
+      "step": 2547200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.286348005028948e-06,
+      "loss": 3.8383,
+      "step": 2547712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.277962057518428e-06,
+      "loss": 3.8707,
+      "step": 2548224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.269576110007908e-06,
+      "loss": 3.8606,
+      "step": 2548736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.261206541301119e-06,
+      "loss": 3.8485,
+      "step": 2549248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.2528205937906e-06,
+      "loss": 3.8436,
+      "step": 2549760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.244434646280079e-06,
+      "loss": 3.8644,
+      "step": 2550272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.236048698769559e-06,
+      "loss": 3.8333,
+      "step": 2550784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.227662751259039e-06,
+      "loss": 3.8466,
+      "step": 2551296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.219276803748519e-06,
+      "loss": 3.8636,
+      "step": 2551808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.210890856237999e-06,
+      "loss": 3.8602,
+      "step": 2552320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.202504908727479e-06,
+      "loss": 3.8525,
+      "step": 2552832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.194118961216959e-06,
+      "loss": 3.8387,
+      "step": 2553344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.185733013706439e-06,
+      "loss": 3.8476,
+      "step": 2553856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.177347066195919e-06,
+      "loss": 3.846,
+      "step": 2554368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.168977497489131e-06,
+      "loss": 3.8599,
+      "step": 2554880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.16059154997861e-06,
+      "loss": 3.8486,
+      "step": 2555392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.15222198127182e-06,
+      "loss": 3.8546,
+      "step": 2555904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.143836033761302e-06,
+      "loss": 3.8576,
+      "step": 2556416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.135450086250782e-06,
+      "loss": 3.8633,
+      "step": 2556928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.12706413874026e-06,
+      "loss": 3.8489,
+      "step": 2557440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.118694570033472e-06,
+      "loss": 3.8553,
+      "step": 2557952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.110308622522952e-06,
+      "loss": 3.8576,
+      "step": 2558464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.10192267501243e-06,
+      "loss": 3.8562,
+      "step": 2558976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.093536727501912e-06,
+      "loss": 3.8522,
+      "step": 2559488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.085150779991392e-06,
+      "loss": 3.8512,
+      "step": 2560000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.07676483248087e-06,
+      "loss": 3.8568,
+      "step": 2560512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.068395263774083e-06,
+      "loss": 3.8553,
+      "step": 2561024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.060009316263563e-06,
+      "loss": 3.8425,
+      "step": 2561536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.051639747556774e-06,
+      "loss": 3.8479,
+      "step": 2562048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.043253800046254e-06,
+      "loss": 3.853,
+      "step": 2562560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.034867852535734e-06,
+      "loss": 3.858,
+      "step": 2563072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.026481905025214e-06,
+      "loss": 3.8536,
+      "step": 2563584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.018095957514694e-06,
+      "loss": 3.8523,
+      "step": 2564096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.009710010004174e-06,
+      "loss": 3.847,
+      "step": 2564608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.001324062493654e-06,
+      "loss": 3.8426,
+      "step": 2565120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.992938114983133e-06,
+      "loss": 3.8459,
+      "step": 2565632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.984552167472613e-06,
+      "loss": 3.8562,
+      "step": 2566144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.976166219962093e-06,
+      "loss": 3.853,
+      "step": 2566656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.967780272451573e-06,
+      "loss": 3.8608,
+      "step": 2567168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.959394324941052e-06,
+      "loss": 3.8587,
+      "step": 2567680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.951008377430533e-06,
+      "loss": 3.847,
+      "step": 2568192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.942638808723744e-06,
+      "loss": 3.8486,
+      "step": 2568704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.934269240016957e-06,
+      "loss": 3.8583,
+      "step": 2569216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.925883292506435e-06,
+      "loss": 3.8606,
+      "step": 2569728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.917497344995915e-06,
+      "loss": 3.8601,
+      "step": 2570240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.909111397485396e-06,
+      "loss": 3.8539,
+      "step": 2570752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.900725449974876e-06,
+      "loss": 3.8459,
+      "step": 2571264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.892339502464355e-06,
+      "loss": 3.8662,
+      "step": 2571776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.883953554953835e-06,
+      "loss": 3.8402,
+      "step": 2572288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.875567607443316e-06,
+      "loss": 3.855,
+      "step": 2572800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.867214417540258e-06,
+      "loss": 3.8537,
+      "step": 2573312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.858828470029738e-06,
+      "loss": 3.8508,
+      "step": 2573824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.850458901322949e-06,
+      "loss": 3.8443,
+      "step": 2574336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.842072953812429e-06,
+      "loss": 3.8528,
+      "step": 2574848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.833687006301909e-06,
+      "loss": 3.8449,
+      "step": 2575360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.825301058791388e-06,
+      "loss": 3.8559,
+      "step": 2575872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.816915111280868e-06,
+      "loss": 3.8521,
+      "step": 2576384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.808529163770348e-06,
+      "loss": 3.8476,
+      "step": 2576896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.800143216259828e-06,
+      "loss": 3.8673,
+      "step": 2577408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.791757268749308e-06,
+      "loss": 3.8684,
+      "step": 2577920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.783371321238788e-06,
+      "loss": 3.8562,
+      "step": 2578432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.775001752531999e-06,
+      "loss": 3.8511,
+      "step": 2578944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.766615805021479e-06,
+      "loss": 3.8482,
+      "step": 2579456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.758229857510959e-06,
+      "loss": 3.8514,
+      "step": 2579968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.74984391000044e-06,
+      "loss": 3.8488,
+      "step": 2580480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.741457962489919e-06,
+      "loss": 3.8596,
+      "step": 2580992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.733088393783131e-06,
+      "loss": 3.8462,
+      "step": 2581504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.724702446272611e-06,
+      "loss": 3.8554,
+      "step": 2582016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.71631649876209e-06,
+      "loss": 3.854,
+      "step": 2582528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.70793055125157e-06,
+      "loss": 3.8585,
+      "step": 2583040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.699544603741051e-06,
+      "loss": 3.8444,
+      "step": 2583552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.691158656230531e-06,
+      "loss": 3.8521,
+      "step": 2584064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.68278908752374e-06,
+      "loss": 3.8497,
+      "step": 2584576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.674419518816953e-06,
+      "loss": 3.8585,
+      "step": 2585088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.666033571306433e-06,
+      "loss": 3.8659,
+      "step": 2585600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.657647623795913e-06,
+      "loss": 3.8637,
+      "step": 2586112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.649261676285393e-06,
+      "loss": 3.8478,
+      "step": 2586624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.640875728774873e-06,
+      "loss": 3.8576,
+      "step": 2587136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.632489781264352e-06,
+      "loss": 3.8577,
+      "step": 2587648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.624103833753832e-06,
+      "loss": 3.8544,
+      "step": 2588160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.615717886243312e-06,
+      "loss": 3.8543,
+      "step": 2588672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.607331938732792e-06,
+      "loss": 3.8628,
+      "step": 2589184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.5989459912222715e-06,
+      "loss": 3.8515,
+      "step": 2589696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.590592801319214e-06,
+      "loss": 3.8729,
+      "step": 2590208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.582206853808695e-06,
+      "loss": 3.8464,
+      "step": 2590720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.573820906298175e-06,
+      "loss": 3.8636,
+      "step": 2591232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.565434958787654e-06,
+      "loss": 3.8531,
+      "step": 2591744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.557049011277134e-06,
+      "loss": 3.8496,
+      "step": 2592256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.5486630637666146e-06,
+      "loss": 3.8572,
+      "step": 2592768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.5402771162560945e-06,
+      "loss": 3.8467,
+      "step": 2593280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.531891168745574e-06,
+      "loss": 3.85,
+      "step": 2593792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.523521600038785e-06,
+      "loss": 3.8553,
+      "step": 2594304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.515135652528266e-06,
+      "loss": 3.8496,
+      "step": 2594816
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.982196569442749,
+      "eval_runtime": 307.3257,
+      "eval_samples_per_second": 1241.65,
+      "eval_steps_per_second": 38.802,
+      "step": 2594864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.506749705017744e-06,
+      "loss": 3.8539,
+      "step": 2595328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.498363757507225e-06,
+      "loss": 3.8501,
+      "step": 2595840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.489977809996705e-06,
+      "loss": 3.8623,
+      "step": 2596352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.481591862486184e-06,
+      "loss": 3.8557,
+      "step": 2596864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.473205914975664e-06,
+      "loss": 3.872,
+      "step": 2597376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.464819967465145e-06,
+      "loss": 3.8503,
+      "step": 2597888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.456450398758357e-06,
+      "loss": 3.8529,
+      "step": 2598400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.448064451247836e-06,
+      "loss": 3.8441,
+      "step": 2598912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.439678503737316e-06,
+      "loss": 3.8609,
+      "step": 2599424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.431308935030527e-06,
+      "loss": 3.8537,
+      "step": 2599936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.422922987520007e-06,
+      "loss": 3.8508,
+      "step": 2600448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.414553418813218e-06,
+      "loss": 3.8639,
+      "step": 2600960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.40618385010643e-06,
+      "loss": 3.8452,
+      "step": 2601472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.39779790259591e-06,
+      "loss": 3.8571,
+      "step": 2601984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.3894119550853904e-06,
+      "loss": 3.8411,
+      "step": 2602496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.381026007574869e-06,
+      "loss": 3.8485,
+      "step": 2603008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.3726400600643495e-06,
+      "loss": 3.8515,
+      "step": 2603520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.3642541125538295e-06,
+      "loss": 3.8534,
+      "step": 2604032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.3558681650433086e-06,
+      "loss": 3.8514,
+      "step": 2604544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.347482217532789e-06,
+      "loss": 3.8743,
+      "step": 2605056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.339096270022269e-06,
+      "loss": 3.8493,
+      "step": 2605568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.3307103225117484e-06,
+      "loss": 3.8565,
+      "step": 2606080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.322324375001228e-06,
+      "loss": 3.8553,
+      "step": 2606592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.313938427490709e-06,
+      "loss": 3.8629,
+      "step": 2607104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.305552479980189e-06,
+      "loss": 3.8521,
+      "step": 2607616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.297166532469668e-06,
+      "loss": 3.8614,
+      "step": 2608128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.288780584959148e-06,
+      "loss": 3.8428,
+      "step": 2608640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.280394637448629e-06,
+      "loss": 3.8525,
+      "step": 2609152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.2720414475455715e-06,
+      "loss": 3.8428,
+      "step": 2609664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.263655500035051e-06,
+      "loss": 3.8575,
+      "step": 2610176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.2552695525245306e-06,
+      "loss": 3.8564,
+      "step": 2610688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.246899983817742e-06,
+      "loss": 3.8579,
+      "step": 2611200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.238514036307221e-06,
+      "loss": 3.8662,
+      "step": 2611712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.230128088796702e-06,
+      "loss": 3.8514,
+      "step": 2612224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.221758520089914e-06,
+      "loss": 3.8601,
+      "step": 2612736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.213372572579394e-06,
+      "loss": 3.8577,
+      "step": 2613248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.204986625068873e-06,
+      "loss": 3.8507,
+      "step": 2613760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.196600677558353e-06,
+      "loss": 3.8459,
+      "step": 2614272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.1882147300478336e-06,
+      "loss": 3.8409,
+      "step": 2614784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.179828782537313e-06,
+      "loss": 3.8474,
+      "step": 2615296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.171442835026793e-06,
+      "loss": 3.8501,
+      "step": 2615808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.163056887516273e-06,
+      "loss": 3.861,
+      "step": 2616320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.154670940005753e-06,
+      "loss": 3.8537,
+      "step": 2616832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.1462849924952325e-06,
+      "loss": 3.8454,
+      "step": 2617344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.1378990449847124e-06,
+      "loss": 3.8584,
+      "step": 2617856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.129513097474192e-06,
+      "loss": 3.85,
+      "step": 2618368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.1211271499636715e-06,
+      "loss": 3.851,
+      "step": 2618880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.112741202453152e-06,
+      "loss": 3.8421,
+      "step": 2619392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.104355254942632e-06,
+      "loss": 3.8413,
+      "step": 2619904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.095969307432111e-06,
+      "loss": 3.8665,
+      "step": 2620416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.087599738725323e-06,
+      "loss": 3.8619,
+      "step": 2620928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.079213791214803e-06,
+      "loss": 3.8471,
+      "step": 2621440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.070844222508015e-06,
+      "loss": 3.8501,
+      "step": 2621952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.062458274997494e-06,
+      "loss": 3.8601,
+      "step": 2622464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.0540723274869745e-06,
+      "loss": 3.8388,
+      "step": 2622976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.0456863799764545e-06,
+      "loss": 3.8585,
+      "step": 2623488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.0373004324659344e-06,
+      "loss": 3.8358,
+      "step": 2624000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.0289144849554135e-06,
+      "loss": 3.8653,
+      "step": 2624512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.020544916248625e-06,
+      "loss": 3.8608,
+      "step": 2625024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.012158968738106e-06,
+      "loss": 3.847,
+      "step": 2625536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.003773021227584e-06,
+      "loss": 3.8392,
+      "step": 2626048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.995387073717065e-06,
+      "loss": 3.8632,
+      "step": 2626560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.987001126206545e-06,
+      "loss": 3.829,
+      "step": 2627072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.978615178696026e-06,
+      "loss": 3.8458,
+      "step": 2627584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.970229231185504e-06,
+      "loss": 3.8606,
+      "step": 2628096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.961843283674985e-06,
+      "loss": 3.8562,
+      "step": 2628608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.953457336164465e-06,
+      "loss": 3.849,
+      "step": 2629120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.945071388653944e-06,
+      "loss": 3.8384,
+      "step": 2629632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.936685441143424e-06,
+      "loss": 3.843,
+      "step": 2630144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.928299493632905e-06,
+      "loss": 3.8429,
+      "step": 2630656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.919929924926115e-06,
+      "loss": 3.8543,
+      "step": 2631168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.911560356219326e-06,
+      "loss": 3.8453,
+      "step": 2631680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.903174408708807e-06,
+      "loss": 3.8552,
+      "step": 2632192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.894788461198287e-06,
+      "loss": 3.8562,
+      "step": 2632704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.886402513687766e-06,
+      "loss": 3.8593,
+      "step": 2633216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.878016566177246e-06,
+      "loss": 3.8454,
+      "step": 2633728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.869630618666727e-06,
+      "loss": 3.8516,
+      "step": 2634240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.861244671156206e-06,
+      "loss": 3.8564,
+      "step": 2634752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.852858723645686e-06,
+      "loss": 3.8499,
+      "step": 2635264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.844489154938898e-06,
+      "loss": 3.8502,
+      "step": 2635776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.8361032074283776e-06,
+      "loss": 3.8513,
+      "step": 2636288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.827717259917857e-06,
+      "loss": 3.854,
+      "step": 2636800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.819347691211068e-06,
+      "loss": 3.8529,
+      "step": 2637312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.810961743700549e-06,
+      "loss": 3.839,
+      "step": 2637824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.802575796190029e-06,
+      "loss": 3.8474,
+      "step": 2638336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.794189848679508e-06,
+      "loss": 3.8488,
+      "step": 2638848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.78582027997272e-06,
+      "loss": 3.8562,
+      "step": 2639360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.7774343324622e-06,
+      "loss": 3.8515,
+      "step": 2639872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.769048384951679e-06,
+      "loss": 3.8483,
+      "step": 2640384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.760662437441159e-06,
+      "loss": 3.8454,
+      "step": 2640896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.75227648993064e-06,
+      "loss": 3.8392,
+      "step": 2641408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.743906921223851e-06,
+      "loss": 3.8428,
+      "step": 2641920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.735537352517061e-06,
+      "loss": 3.8501,
+      "step": 2642432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.727151405006542e-06,
+      "loss": 3.8579,
+      "step": 2642944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.718765457496022e-06,
+      "loss": 3.8603,
+      "step": 2643456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.710379509985503e-06,
+      "loss": 3.8521,
+      "step": 2643968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.702009941278713e-06,
+      "loss": 3.8477,
+      "step": 2644480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.693623993768193e-06,
+      "loss": 3.8458,
+      "step": 2644992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.6852380462576735e-06,
+      "loss": 3.8518,
+      "step": 2645504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.676852098747153e-06,
+      "loss": 3.859,
+      "step": 2646016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.6684661512366325e-06,
+      "loss": 3.8552,
+      "step": 2646528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.6600802037261125e-06,
+      "loss": 3.8511,
+      "step": 2647040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.651694256215593e-06,
+      "loss": 3.8438,
+      "step": 2647552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.643308308705072e-06,
+      "loss": 3.8613,
+      "step": 2648064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.634922361194552e-06,
+      "loss": 3.8402,
+      "step": 2648576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.626536413684032e-06,
+      "loss": 3.8528,
+      "step": 2649088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.618166844977243e-06,
+      "loss": 3.8486,
+      "step": 2649600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.609780897466723e-06,
+      "loss": 3.8495,
+      "step": 2650112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.601394949956204e-06,
+      "loss": 3.8414,
+      "step": 2650624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.593009002445684e-06,
+      "loss": 3.8534,
+      "step": 2651136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.584623054935163e-06,
+      "loss": 3.8404,
+      "step": 2651648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.576237107424643e-06,
+      "loss": 3.8503,
+      "step": 2652160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.567851159914124e-06,
+      "loss": 3.8551,
+      "step": 2652672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.559481591207334e-06,
+      "loss": 3.8398,
+      "step": 2653184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.551112022500545e-06,
+      "loss": 3.8648,
+      "step": 2653696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.542726074990026e-06,
+      "loss": 3.8646,
+      "step": 2654208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.534340127479506e-06,
+      "loss": 3.8566,
+      "step": 2654720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.525954179968985e-06,
+      "loss": 3.8454,
+      "step": 2655232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.517568232458465e-06,
+      "loss": 3.846,
+      "step": 2655744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.509182284947946e-06,
+      "loss": 3.8504,
+      "step": 2656256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.500796337437424e-06,
+      "loss": 3.8429,
+      "step": 2656768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.492426768730636e-06,
+      "loss": 3.8599,
+      "step": 2657280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.484040821220117e-06,
+      "loss": 3.8402,
+      "step": 2657792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.475671252513328e-06,
+      "loss": 3.854,
+      "step": 2658304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.467285305002807e-06,
+      "loss": 3.8534,
+      "step": 2658816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.458899357492287e-06,
+      "loss": 3.8513,
+      "step": 2659328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.450529788785499e-06,
+      "loss": 3.8423,
+      "step": 2659840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.44214384127498e-06,
+      "loss": 3.8493,
+      "step": 2660352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.433757893764458e-06,
+      "loss": 3.8506,
+      "step": 2660864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.425371946253939e-06,
+      "loss": 3.8553,
+      "step": 2661376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.416985998743419e-06,
+      "loss": 3.8567,
+      "step": 2661888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.408600051232898e-06,
+      "loss": 3.8636,
+      "step": 2662400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.400214103722378e-06,
+      "loss": 3.8438,
+      "step": 2662912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.391828156211859e-06,
+      "loss": 3.8571,
+      "step": 2663424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.3834422087013386e-06,
+      "loss": 3.856,
+      "step": 2663936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.375056261190818e-06,
+      "loss": 3.8498,
+      "step": 2664448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.366686692484029e-06,
+      "loss": 3.8504,
+      "step": 2664960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.358300744973509e-06,
+      "loss": 3.8615,
+      "step": 2665472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.349914797462988e-06,
+      "loss": 3.8522,
+      "step": 2665984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.341528849952469e-06,
+      "loss": 3.866,
+      "step": 2666496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.333142902441949e-06,
+      "loss": 3.8481,
+      "step": 2667008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.324756954931428e-06,
+      "loss": 3.8581,
+      "step": 2667520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.316371007420908e-06,
+      "loss": 3.8522,
+      "step": 2668032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.307985059910389e-06,
+      "loss": 3.8478,
+      "step": 2668544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.299615491203601e-06,
+      "loss": 3.8518,
+      "step": 2669056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.291229543693079e-06,
+      "loss": 3.8464,
+      "step": 2669568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.28284359618256e-06,
+      "loss": 3.847,
+      "step": 2670080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.274490406279503e-06,
+      "loss": 3.8508,
+      "step": 2670592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.266104458768983e-06,
+      "loss": 3.8468,
+      "step": 2671104
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.981346607208252,
+      "eval_runtime": 304.0264,
+      "eval_samples_per_second": 1255.124,
+      "eval_steps_per_second": 39.224,
+      "step": 2671184
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.0823006598525196e+18,
+  "trial_name": null,
+  "trial_params": null
+}