diff --git "a/checkpoint-2213268/trainer_state.json" "b/checkpoint-2213268/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2213268/trainer_state.json"
@@ -0,0 +1,26189 @@
+{
+  "best_metric": 3.977329730987549,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/lstm/0/checkpoints/checkpoint-2213268",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 2213268,
+  "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": 7.5643,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0552,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9853,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9546,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8345,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.6849,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.578,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5051,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4266,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3936,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3397,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2845,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2212,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1692,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987424354494966e-05,
+      "loss": 6.1253,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986585759743914e-05,
+      "loss": 6.0774,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985747164992862e-05,
+      "loss": 6.0212,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98490857024181e-05,
+      "loss": 5.9826,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984069975490758e-05,
+      "loss": 5.9437,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.8964,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8603,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8387,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.8015,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7798,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.753,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7306,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.7024,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6697,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6631,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6306,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6149,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5898,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5863,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5553,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5464,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5332,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.5146,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.5046,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4821,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.458,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4517,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4245,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.4377,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.4104,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.4062,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.39,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3878,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3812,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3515,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3399,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3333,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3401,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.3164,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.3222,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2935,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2899,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2768,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2704,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2681,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2477,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2419,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2455,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2254,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.2229,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2119,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.2113,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1964,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1989,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1858,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1748,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.178,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1629,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1712,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1684,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1604,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1398,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1243,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.1169,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.1225,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 5.1249,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 5.1188,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.0935,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0878,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0967,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0827,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0815,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.0732,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0747,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0627,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0555,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0409,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0549,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0393,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.023,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0225,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.0346,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 5.0247,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 5.0148,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 5.0016,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 5.0071,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9867,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 5.0018,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 5.0018,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9766,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9698,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9484,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9585,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9589,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.9543,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9548,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.9378,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9401,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.9402,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9417,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9285,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.917,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9183,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.9209,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.9158,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9127,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.9006,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8934,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8861,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8922,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8892,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8909,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8653,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8795,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8787,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8649,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8806,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8592,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8732,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8616,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8408,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.84,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8401,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884314871364151e-05,
+      "loss": 4.842,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.847,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.8436,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.8373,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.8266,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8152,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.8153,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8211,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.8283,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8185,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.802,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.8015,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.766555309295654,
+      "eval_runtime": 557.3806,
+      "eval_samples_per_second": 684.615,
+      "eval_steps_per_second": 21.395,
+      "step": 76319
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.8093,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.8013,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7935,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7915,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7826,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7666,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7752,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7838,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.767,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7745,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.777,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7775,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.765,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7614,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.7594,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.7504,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.7528,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7522,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7511,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.75,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857489666612726e-05,
+      "loss": 4.7362,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.7489,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.7349,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.744,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.733,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7329,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7341,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7159,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7296,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.7102,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.7158,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.7127,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.72,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846589572729423e-05,
+      "loss": 4.7078,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.7009,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.718,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.7027,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.7103,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6965,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6849,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.6876,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.6795,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.6936,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838206900979649e-05,
+      "loss": 4.6825,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6895,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.6801,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.6877,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.6846,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834015565104763e-05,
+      "loss": 4.6593,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833176970353711e-05,
+      "loss": 4.6648,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.663,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6745,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6529,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6769,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6504,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828147039727771e-05,
+      "loss": 4.6505,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827308444976719e-05,
+      "loss": 4.653,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826469850225667e-05,
+      "loss": 4.6485,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.6501,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247942986039366e-05,
+      "loss": 4.6425,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.6398,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.6484,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222801522311535e-05,
+      "loss": 4.6354,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214415574801015e-05,
+      "loss": 4.6349,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6328,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.638,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.6225,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.6415,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172502216052144e-05,
+      "loss": 4.6283,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6265,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.6312,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.6191,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81389748048138e-05,
+      "loss": 4.6362,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.6398,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812220290979276e-05,
+      "loss": 4.6348,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.6161,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6095,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.6044,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808867549855441e-05,
+      "loss": 4.6155,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808028955104389e-05,
+      "loss": 4.6255,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807190360353337e-05,
+      "loss": 4.6182,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806351765602285e-05,
+      "loss": 4.6047,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805513170851233e-05,
+      "loss": 4.5933,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804674576100181e-05,
+      "loss": 4.6112,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803835981349129e-05,
+      "loss": 4.5969,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8029990244784504e-05,
+      "loss": 4.6,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.6018,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.6099,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800484878105667e-05,
+      "loss": 4.5978,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799646283354615e-05,
+      "loss": 4.5951,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.5844,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5974,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5947,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.5723,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795454947479728e-05,
+      "loss": 4.5805,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794616352728676e-05,
+      "loss": 4.5964,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.5966,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5821,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5781,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5875,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.5677,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789586422102737e-05,
+      "loss": 4.5834,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788747827351685e-05,
+      "loss": 4.5865,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787909232600633e-05,
+      "loss": 4.573,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787070637849581e-05,
+      "loss": 4.5731,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786232043098529e-05,
+      "loss": 4.5513,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785393448347477e-05,
+      "loss": 4.5566,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784556491476798e-05,
+      "loss": 4.563,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783717896725746e-05,
+      "loss": 4.5636,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782879301974694e-05,
+      "loss": 4.568,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782040707223642e-05,
+      "loss": 4.5561,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812021124725906e-05,
+      "loss": 4.554,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803651556019115e-05,
+      "loss": 4.5605,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795265608508595e-05,
+      "loss": 4.5666,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786879660998075e-05,
+      "loss": 4.5541,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778493713487555e-05,
+      "loss": 4.5549,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770124144780764e-05,
+      "loss": 4.5507,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761738197270244e-05,
+      "loss": 4.5556,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753352249759724e-05,
+      "loss": 4.5553,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7744966302249204e-05,
+      "loss": 4.5483,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736580354738684e-05,
+      "loss": 4.5469,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728194407228164e-05,
+      "loss": 4.5414,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719808459717644e-05,
+      "loss": 4.536,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711422512207124e-05,
+      "loss": 4.5419,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703036564696604e-05,
+      "loss": 4.5452,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694650617186084e-05,
+      "loss": 4.5475,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686264669675564e-05,
+      "loss": 4.5289,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7677878722165044e-05,
+      "loss": 4.5387,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766950915345826e-05,
+      "loss": 4.5503,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766113958475147e-05,
+      "loss": 4.5321,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765277001604468e-05,
+      "loss": 4.5458,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764438406853416e-05,
+      "loss": 4.5315,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763599812102364e-05,
+      "loss": 4.5482,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762761217351312e-05,
+      "loss": 4.5357,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76192262260026e-05,
+      "loss": 4.5208,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761084027849208e-05,
+      "loss": 4.5233,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760245433098156e-05,
+      "loss": 4.52,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759406838347104e-05,
+      "loss": 4.5297,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758568243596052e-05,
+      "loss": 4.5318,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757729648845e-05,
+      "loss": 4.533,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756891054093948e-05,
+      "loss": 4.5285,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756052459342896e-05,
+      "loss": 4.5203,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75521714035259e-05,
+      "loss": 4.5146,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543785456015375e-05,
+      "loss": 4.5121,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535399508504855e-05,
+      "loss": 4.5211,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527013560994335e-05,
+      "loss": 4.534,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.5185,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751025804477703e-05,
+      "loss": 4.5094,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750187209726651e-05,
+      "loss": 4.5084,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.478529930114746,
+      "eval_runtime": 543.7518,
+      "eval_samples_per_second": 701.774,
+      "eval_steps_per_second": 21.931,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749348614975599e-05,
+      "loss": 4.52,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748510020224547e-05,
+      "loss": 4.5158,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747671425473495e-05,
+      "loss": 4.5033,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746832830722443e-05,
+      "loss": 4.5046,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745994235971391e-05,
+      "loss": 4.5017,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745155641220339e-05,
+      "loss": 4.4877,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744317046469287e-05,
+      "loss": 4.4971,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743478451718235e-05,
+      "loss": 4.5087,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742639856967183e-05,
+      "loss": 4.4885,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741801262216131e-05,
+      "loss": 4.4977,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740962667465079e-05,
+      "loss": 4.508,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401240727140264e-05,
+      "loss": 4.5063,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739285477962975e-05,
+      "loss": 4.4939,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384485210922966e-05,
+      "loss": 4.4944,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376099263412446e-05,
+      "loss": 4.4918,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4882,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.4858,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4919,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4929,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.4909,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.482,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4929,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4814,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4879,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4834,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4818,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.4857,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267098324579415e-05,
+      "loss": 4.4701,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258712377068895e-05,
+      "loss": 4.4844,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4708,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.4733,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4729,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4773,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216799018320024e-05,
+      "loss": 4.4704,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208413070809504e-05,
+      "loss": 4.462,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4812,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719167393339592e-05,
+      "loss": 4.4683,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71832879858854e-05,
+      "loss": 4.4814,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717490203837488e-05,
+      "loss": 4.4684,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716651609086437e-05,
+      "loss": 4.4531,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715813014335385e-05,
+      "loss": 4.4543,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.4557,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714135824833281e-05,
+      "loss": 4.4725,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713297230082228e-05,
+      "loss": 4.4508,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712458635331176e-05,
+      "loss": 4.4652,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711620040580124e-05,
+      "loss": 4.4605,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4604,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.4656,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709105894207341e-05,
+      "loss": 4.4401,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708267299456289e-05,
+      "loss": 4.4456,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707428704705237e-05,
+      "loss": 4.4395,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706590109954185e-05,
+      "loss": 4.4607,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705751515203134e-05,
+      "loss": 4.4383,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704912920452082e-05,
+      "loss": 4.4621,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040776014617755e-05,
+      "loss": 4.4366,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032390067107235e-05,
+      "loss": 4.4376,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024004119596715e-05,
+      "loss": 4.4398,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4352,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4437,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998862655868884e-05,
+      "loss": 4.4377,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990476708358364e-05,
+      "loss": 4.4271,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4433,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4305,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965318865826804e-05,
+      "loss": 4.4354,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695693291831629e-05,
+      "loss": 4.429,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694854697080577e-05,
+      "loss": 4.4371,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694016102329525e-05,
+      "loss": 4.4173,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693179145458846e-05,
+      "loss": 4.4433,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4289,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.4312,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.4319,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4244,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688989447464332e-05,
+      "loss": 4.4407,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4455,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4431,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4278,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856350684601245e-05,
+      "loss": 4.418,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847964737090725e-05,
+      "loss": 4.409,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839578789580204e-05,
+      "loss": 4.43,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831192842069684e-05,
+      "loss": 4.4356,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822806894559164e-05,
+      "loss": 4.4316,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814437325852373e-05,
+      "loss": 4.4178,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806051378341853e-05,
+      "loss": 4.4075,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797665430831333e-05,
+      "loss": 4.422,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678927948332081e-05,
+      "loss": 4.4078,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6780893535810287e-05,
+      "loss": 4.4144,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67725239671035e-05,
+      "loss": 4.4229,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676413801959298e-05,
+      "loss": 4.4217,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675575207208246e-05,
+      "loss": 4.4163,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674736612457194e-05,
+      "loss": 4.4181,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673898017706142e-05,
+      "loss": 4.4026,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67305942295509e-05,
+      "loss": 4.4154,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672220828204038e-05,
+      "loss": 4.4133,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671383871333359e-05,
+      "loss": 4.3977,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670545276582307e-05,
+      "loss": 4.4032,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669706681831255e-05,
+      "loss": 4.4181,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668868087080203e-05,
+      "loss": 4.421,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668031130209524e-05,
+      "loss": 4.405,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667192535458472e-05,
+      "loss": 4.407,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663555785877936e-05,
+      "loss": 4.4132,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655169838367416e-05,
+      "loss": 4.395,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6646783890856896e-05,
+      "loss": 4.4109,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.4157,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630011995835856e-05,
+      "loss": 4.4006,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621626048325336e-05,
+      "loss": 4.4074,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613240100814816e-05,
+      "loss": 4.3833,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604854153304296e-05,
+      "loss": 4.3843,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596468205793776e-05,
+      "loss": 4.3952,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588082258283256e-05,
+      "loss": 4.3967,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579696310772736e-05,
+      "loss": 4.4036,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571310363262216e-05,
+      "loss": 4.3882,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562940794555425e-05,
+      "loss": 4.3874,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554554847044905e-05,
+      "loss": 4.3917,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546168899534385e-05,
+      "loss": 4.4029,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537782952023865e-05,
+      "loss": 4.3942,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652941338331708e-05,
+      "loss": 4.3981,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652102743580656e-05,
+      "loss": 4.3871,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651264148829604e-05,
+      "loss": 4.3884,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650427191958925e-05,
+      "loss": 4.4018,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649588597207873e-05,
+      "loss": 4.3869,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648750002456821e-05,
+      "loss": 4.389,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647911407705769e-05,
+      "loss": 4.3856,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647072812954717e-05,
+      "loss": 4.3773,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646234218203665e-05,
+      "loss": 4.3845,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645397261332986e-05,
+      "loss": 4.3852,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.3939,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.3715,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64288147707983e-05,
+      "loss": 4.3803,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420445202091514e-05,
+      "loss": 4.3987,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412059254580994e-05,
+      "loss": 4.3778,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403673307070474e-05,
+      "loss": 4.3925,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639528735955995e-05,
+      "loss": 4.3805,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638690141204943e-05,
+      "loss": 4.3958,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637851546453891e-05,
+      "loss": 4.3829,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637012951702839e-05,
+      "loss": 4.3735,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.3732,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353374000811076e-05,
+      "loss": 4.3673,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6344988053300556e-05,
+      "loss": 4.3786,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336602105790036e-05,
+      "loss": 4.3877,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632823253708325e-05,
+      "loss": 4.384,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631984658957273e-05,
+      "loss": 4.3789,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631146064206221e-05,
+      "loss": 4.3715,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630307469455169e-05,
+      "loss": 4.3695,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629468874704117e-05,
+      "loss": 4.3651,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628631917833438e-05,
+      "loss": 4.375,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627793323082386e-05,
+      "loss": 4.3848,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626954728331334e-05,
+      "loss": 4.3745,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626116133580282e-05,
+      "loss": 4.3671,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62527753882923e-05,
+      "loss": 4.3627,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.342203617095947,
+      "eval_runtime": 526.193,
+      "eval_samples_per_second": 725.192,
+      "eval_steps_per_second": 22.663,
+      "step": 228957
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624440581958551e-05,
+      "loss": 4.3743,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.3768,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.3617,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219247977053956e-05,
+      "loss": 4.3602,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210862029543436e-05,
+      "loss": 4.3632,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.347,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194106513326125e-05,
+      "loss": 4.3557,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185720565815605e-05,
+      "loss": 4.3658,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177334618305085e-05,
+      "loss": 4.3542,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.3616,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160579102087774e-05,
+      "loss": 4.3678,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152193154577254e-05,
+      "loss": 4.3642,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6143807207066734e-05,
+      "loss": 4.3569,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135421259556214e-05,
+      "loss": 4.3619,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127035312045694e-05,
+      "loss": 4.3539,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118649364535174e-05,
+      "loss": 4.3537,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110263417024654e-05,
+      "loss": 4.3485,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610187746951414e-05,
+      "loss": 4.3558,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.3587,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3603,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3521,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3611,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605996411076526e-05,
+      "loss": 4.3507,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605157816325474e-05,
+      "loss": 4.3508,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604320859454796e-05,
+      "loss": 4.3527,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603482264703743e-05,
+      "loss": 4.3487,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602643669952691e-05,
+      "loss": 4.354,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601805075201639e-05,
+      "loss": 4.3411,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600966480450588e-05,
+      "loss": 4.3555,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600127885699536e-05,
+      "loss": 4.3411,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599289290948484e-05,
+      "loss": 4.3435,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598452334077805e-05,
+      "loss": 4.3456,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3537,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.339,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595936549824649e-05,
+      "loss": 4.3414,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5950995929539696e-05,
+      "loss": 4.3498,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942609982029176e-05,
+      "loss": 4.3414,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934224034518656e-05,
+      "loss": 4.3604,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5925838087008136e-05,
+      "loss": 4.3416,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917468518301345e-05,
+      "loss": 4.3289,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590908257079083e-05,
+      "loss": 4.3315,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590069662328031e-05,
+      "loss": 4.3293,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589231067576979e-05,
+      "loss": 4.349,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588392472825927e-05,
+      "loss": 4.3304,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587553878074875e-05,
+      "loss": 4.3413,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586715283323823e-05,
+      "loss": 4.3383,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585876688572771e-05,
+      "loss": 4.3381,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585039731702092e-05,
+      "loss": 4.3442,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58420113695104e-05,
+      "loss": 4.3217,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583362542199988e-05,
+      "loss": 4.3214,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582523947448936e-05,
+      "loss": 4.326,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581685352697884e-05,
+      "loss": 4.3394,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580846757946832e-05,
+      "loss": 4.318,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580009801076153e-05,
+      "loss": 4.3396,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791712063251017e-05,
+      "loss": 4.3201,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783326115740497e-05,
+      "loss": 4.3159,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577494016822997e-05,
+      "loss": 4.325,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576655422071945e-05,
+      "loss": 4.3134,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575816827320893e-05,
+      "loss": 4.3312,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574978232569841e-05,
+      "loss": 4.3225,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574139637818789e-05,
+      "loss": 4.3085,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57330268094811e-05,
+      "loss": 4.3255,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572464086197058e-05,
+      "loss": 4.3156,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716271293263794e-05,
+      "loss": 4.3206,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570788534575327e-05,
+      "loss": 4.3112,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699499398242754e-05,
+      "loss": 4.3248,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691113450732234e-05,
+      "loss": 4.3037,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682727503221714e-05,
+      "loss": 4.3274,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674341555711194e-05,
+      "loss": 4.3133,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5665955608200674e-05,
+      "loss": 4.3146,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657569660690154e-05,
+      "loss": 4.3204,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.3143,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564083052327657e-05,
+      "loss": 4.3238,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563244457576605e-05,
+      "loss": 4.3344,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562405862825553e-05,
+      "loss": 4.329,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561568905954874e-05,
+      "loss": 4.3153,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560730311203822e-05,
+      "loss": 4.309,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559891716452771e-05,
+      "loss": 4.3013,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559053121701719e-05,
+      "loss": 4.3175,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3253,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557375932199615e-05,
+      "loss": 4.3212,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556537337448563e-05,
+      "loss": 4.3061,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555698742697511e-05,
+      "loss": 4.2982,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554860147946459e-05,
+      "loss": 4.3179,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554021553195407e-05,
+      "loss": 4.2941,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553184596324728e-05,
+      "loss": 4.3083,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552346001573676e-05,
+      "loss": 4.3112,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.3138,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506704499519446e-05,
+      "loss": 4.3103,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498318552008926e-05,
+      "loss": 4.3143,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489932604498406e-05,
+      "loss": 4.2951,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481546656987886e-05,
+      "loss": 4.3061,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54731770882811e-05,
+      "loss": 4.3043,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546479114077058e-05,
+      "loss": 4.2961,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.2961,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3163,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543963329823902e-05,
+      "loss": 4.3093,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54312473507285e-05,
+      "loss": 4.3066,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542286140321798e-05,
+      "loss": 4.3011,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414475455707455e-05,
+      "loss": 4.309,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406089508196934e-05,
+      "loss": 4.2922,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397703560686414e-05,
+      "loss": 4.3059,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389317613175894e-05,
+      "loss": 4.312,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380931665665374e-05,
+      "loss": 4.292,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537256209695859e-05,
+      "loss": 4.3106,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536417614944807e-05,
+      "loss": 4.2758,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535579020193755e-05,
+      "loss": 4.2827,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534742063323076e-05,
+      "loss": 4.2946,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533903468572024e-05,
+      "loss": 4.2956,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533064873820972e-05,
+      "loss": 4.3044,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53222627906992e-05,
+      "loss": 4.2857,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531387684318868e-05,
+      "loss": 4.2855,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530549089567816e-05,
+      "loss": 4.2907,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529710494816764e-05,
+      "loss": 4.3032,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528871900065712e-05,
+      "loss": 4.2969,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528034943195033e-05,
+      "loss": 4.2987,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527199624204728e-05,
+      "loss": 4.2864,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526361029453675e-05,
+      "loss": 4.2888,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525522434702623e-05,
+      "loss": 4.3051,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524683839951571e-05,
+      "loss": 4.2878,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523845245200519e-05,
+      "loss": 4.2857,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523006650449467e-05,
+      "loss": 4.2901,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.2765,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521329460947363e-05,
+      "loss": 4.2904,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520490866196311e-05,
+      "loss": 4.2876,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519652271445259e-05,
+      "loss": 4.2974,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.2756,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517976719823528e-05,
+      "loss": 4.2811,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517138125072476e-05,
+      "loss": 4.3047,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516301168201798e-05,
+      "loss": 4.2816,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515462573450746e-05,
+      "loss": 4.2914,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514623978699694e-05,
+      "loss": 4.2889,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513785383948642e-05,
+      "loss": 4.2994,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51294678919759e-05,
+      "loss": 4.2897,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512108194446538e-05,
+      "loss": 4.2802,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511269599695486e-05,
+      "loss": 4.2751,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510431004944434e-05,
+      "loss": 4.2733,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509592410193382e-05,
+      "loss": 4.2836,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508753815442329e-05,
+      "loss": 4.2938,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507915220691277e-05,
+      "loss": 4.2887,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507076625940225e-05,
+      "loss": 4.2893,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062396690695466e-05,
+      "loss": 4.2769,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505402712198868e-05,
+      "loss": 4.2783,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504564117447816e-05,
+      "loss": 4.2708,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503725522696764e-05,
+      "loss": 4.2841,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028869279457115e-05,
+      "loss": 4.2861,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020483331946595e-05,
+      "loss": 4.2903,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012097384436075e-05,
+      "loss": 4.2732,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003711436925555e-05,
+      "loss": 4.2719,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.258720874786377,
+      "eval_runtime": 536.684,
+      "eval_samples_per_second": 711.016,
+      "eval_steps_per_second": 22.22,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995325489415035e-05,
+      "loss": 4.2865,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.2824,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.2703,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.2707,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.2713,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.2591,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.2656,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.2703,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.2683,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919851961820354e-05,
+      "loss": 4.2745,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911466014309834e-05,
+      "loss": 4.2781,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903080066799314e-05,
+      "loss": 4.2712,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894694119288794e-05,
+      "loss": 4.2709,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886308171778274e-05,
+      "loss": 4.2733,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877922224267754e-05,
+      "loss": 4.2686,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869536276757234e-05,
+      "loss": 4.2633,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861150329246714e-05,
+      "loss": 4.2596,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4852764381736194e-05,
+      "loss": 4.2727,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48443948130294e-05,
+      "loss": 4.2686,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483600886551888e-05,
+      "loss": 4.2762,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482762291800836e-05,
+      "loss": 4.262,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481923697049785e-05,
+      "loss": 4.2731,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481085102298733e-05,
+      "loss": 4.2678,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480246507547681e-05,
+      "loss": 4.2656,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479407912796629e-05,
+      "loss": 4.2654,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2622,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477733999055271e-05,
+      "loss": 4.2694,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476895404304219e-05,
+      "loss": 4.2538,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476056809553167e-05,
+      "loss": 4.2677,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475218214802115e-05,
+      "loss": 4.2584,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474379620051063e-05,
+      "loss": 4.2548,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473541025300011e-05,
+      "loss": 4.2635,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.266,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471863835797907e-05,
+      "loss": 4.2593,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471025241046855e-05,
+      "loss": 4.2573,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701882841761763e-05,
+      "loss": 4.2612,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693496894251243e-05,
+      "loss": 4.2622,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685110946740717e-05,
+      "loss": 4.2749,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676724999230197e-05,
+      "loss": 4.2582,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668339051719676e-05,
+      "loss": 4.2475,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4659953104209156e-05,
+      "loss": 4.246,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651567156698636e-05,
+      "loss": 4.2493,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643197587991846e-05,
+      "loss": 4.2635,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634811640481325e-05,
+      "loss": 4.249,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626425692970805e-05,
+      "loss": 4.2617,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618039745460285e-05,
+      "loss": 4.2572,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609653797949765e-05,
+      "loss": 4.253,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460128422924298e-05,
+      "loss": 4.2632,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459289828173246e-05,
+      "loss": 4.2429,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458451233422194e-05,
+      "loss": 4.2356,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457612638671142e-05,
+      "loss": 4.2489,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45677404392009e-05,
+      "loss": 4.2547,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455935449169038e-05,
+      "loss": 4.2364,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455096854417986e-05,
+      "loss": 4.2593,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454259897547307e-05,
+      "loss": 4.2417,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453421302796255e-05,
+      "loss": 4.2343,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452582708045203e-05,
+      "loss": 4.2447,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451744113294151e-05,
+      "loss": 4.2341,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450905518543099e-05,
+      "loss": 4.2529,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450066923792047e-05,
+      "loss": 4.2458,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492299669213686e-05,
+      "loss": 4.2278,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483913721703166e-05,
+      "loss": 4.2442,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475527774192646e-05,
+      "loss": 4.2374,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467141826682126e-05,
+      "loss": 4.2468,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458755879171606e-05,
+      "loss": 4.2318,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445036993166108e-05,
+      "loss": 4.2448,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444198398415056e-05,
+      "loss": 4.2272,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443359803664004e-05,
+      "loss": 4.2462,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425228467933254e-05,
+      "loss": 4.2357,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441684252042273e-05,
+      "loss": 4.2355,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440845657291221e-05,
+      "loss": 4.245,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440007062540169e-05,
+      "loss": 4.2363,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391701056694903e-05,
+      "loss": 4.2454,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438331510918438e-05,
+      "loss": 4.2549,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437492916167386e-05,
+      "loss": 4.2546,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436654321416334e-05,
+      "loss": 4.2388,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435815726665282e-05,
+      "loss": 4.2337,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434978769794603e-05,
+      "loss": 4.2246,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434141812923925e-05,
+      "loss": 4.2383,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433303218172873e-05,
+      "loss": 4.2535,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43246462342182e-05,
+      "loss": 4.2425,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431626028670768e-05,
+      "loss": 4.2329,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430787433919716e-05,
+      "loss": 4.2194,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429948839168664e-05,
+      "loss": 4.2464,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429111882297986e-05,
+      "loss": 4.2223,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428274925427307e-05,
+      "loss": 4.2307,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427436330676255e-05,
+      "loss": 4.234,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4265977359252026e-05,
+      "loss": 4.2377,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257591411741506e-05,
+      "loss": 4.2384,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249205464230986e-05,
+      "loss": 4.2395,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240819516720466e-05,
+      "loss": 4.2227,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232433569209946e-05,
+      "loss": 4.2332,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224047621699426e-05,
+      "loss": 4.2322,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215661674188906e-05,
+      "loss": 4.2235,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207275726678386e-05,
+      "loss": 4.2214,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198889779167866e-05,
+      "loss": 4.2437,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190503831657346e-05,
+      "loss": 4.2365,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418213426295056e-05,
+      "loss": 4.2351,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417374831544004e-05,
+      "loss": 4.229,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416536236792952e-05,
+      "loss": 4.2326,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4156976420419e-05,
+      "loss": 4.2217,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414860685171221e-05,
+      "loss": 4.2348,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414022090420169e-05,
+      "loss": 4.2367,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413183495669117e-05,
+      "loss": 4.2226,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412344900918065e-05,
+      "loss": 4.2388,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411506306167013e-05,
+      "loss": 4.2069,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410670987176707e-05,
+      "loss": 4.2133,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409832392425655e-05,
+      "loss": 4.2205,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408993797674603e-05,
+      "loss": 4.224,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081552029235515e-05,
+      "loss": 4.2328,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073166081724995e-05,
+      "loss": 4.22,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064780134214475e-05,
+      "loss": 4.2117,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056394186703955e-05,
+      "loss": 4.2191,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048008239193435e-05,
+      "loss": 4.2314,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039638670486644e-05,
+      "loss": 4.2268,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031252722976124e-05,
+      "loss": 4.2316,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4022866775465604e-05,
+      "loss": 4.2173,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014480827955084e-05,
+      "loss": 4.2199,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400611125924829e-05,
+      "loss": 4.2335,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399772531173777e-05,
+      "loss": 4.2216,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398933936422725e-05,
+      "loss": 4.2153,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398095341671673e-05,
+      "loss": 4.2197,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397256746920621e-05,
+      "loss": 4.2093,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396418152169569e-05,
+      "loss": 4.2193,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395581195298891e-05,
+      "loss": 4.2208,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394742600547839e-05,
+      "loss": 4.2257,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393904005796786e-05,
+      "loss": 4.2089,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393065411045734e-05,
+      "loss": 4.2143,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392226816294682e-05,
+      "loss": 4.2315,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391389859424003e-05,
+      "loss": 4.2178,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390551264672951e-05,
+      "loss": 4.223,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389712669921899e-05,
+      "loss": 4.2237,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388874075170847e-05,
+      "loss": 4.2272,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388035480419795e-05,
+      "loss": 4.226,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387196885668744e-05,
+      "loss": 4.2107,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386358290917692e-05,
+      "loss": 4.2072,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38551969616664e-05,
+      "loss": 4.2114,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846827392959606e-05,
+      "loss": 4.2169,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838441445449086e-05,
+      "loss": 4.2228,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830055497938566e-05,
+      "loss": 4.225,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821669550428046e-05,
+      "loss": 4.2206,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813283602917526e-05,
+      "loss": 4.213,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804897655407006e-05,
+      "loss": 4.2117,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796511707896486e-05,
+      "loss": 4.2034,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788142139189695e-05,
+      "loss": 4.2192,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779756191679175e-05,
+      "loss": 4.2217,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771370244168655e-05,
+      "loss": 4.2246,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3762984296658135e-05,
+      "loss": 4.2045,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375459834914762e-05,
+      "loss": 4.207,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.200766563415527,
+      "eval_runtime": 557.9084,
+      "eval_samples_per_second": 683.967,
+      "eval_steps_per_second": 21.374,
+      "step": 381595
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37462124016371e-05,
+      "loss": 4.2241,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373784283293031e-05,
+      "loss": 4.2131,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372945688541979e-05,
+      "loss": 4.2077,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372107093790927e-05,
+      "loss": 4.202,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371268499039875e-05,
+      "loss": 4.2056,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3704299042888224e-05,
+      "loss": 4.1973,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3695913095377704e-05,
+      "loss": 4.2031,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3687527147867184e-05,
+      "loss": 4.1999,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3679141200356664e-05,
+      "loss": 4.2058,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367078801045361e-05,
+      "loss": 4.2149,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366240206294309e-05,
+      "loss": 4.2086,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654016115432576e-05,
+      "loss": 4.2071,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364563016792205e-05,
+      "loss": 4.2101,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363724422041153e-05,
+      "loss": 4.2095,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362885827290101e-05,
+      "loss": 4.2033,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362047232539049e-05,
+      "loss": 4.199,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361208637787997e-05,
+      "loss": 4.1967,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360370043036945e-05,
+      "loss": 4.2063,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359533086166266e-05,
+      "loss": 4.206,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358694491415214e-05,
+      "loss": 4.2143,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357855896664162e-05,
+      "loss": 4.2026,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35701730191311e-05,
+      "loss": 4.2103,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356178707162058e-05,
+      "loss": 4.2068,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355341750291379e-05,
+      "loss": 4.2041,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545047934207e-05,
+      "loss": 4.202,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353666198669648e-05,
+      "loss": 4.1999,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352827603918596e-05,
+      "loss": 4.2081,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351989009167544e-05,
+      "loss": 4.1899,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351150414416492e-05,
+      "loss": 4.2107,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35031181966544e-05,
+      "loss": 4.1957,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349473224914388e-05,
+      "loss": 4.1915,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348634630163336e-05,
+      "loss": 4.2048,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347796035412284e-05,
+      "loss": 4.2024,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346959078541605e-05,
+      "loss": 4.2026,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346120483790553e-05,
+      "loss": 4.1974,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345283526919875e-05,
+      "loss": 4.197,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344444932168823e-05,
+      "loss": 4.2001,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343606337417771e-05,
+      "loss": 4.2161,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342767742666719e-05,
+      "loss": 4.1984,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341929147915667e-05,
+      "loss": 4.1877,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341090553164615e-05,
+      "loss": 4.1884,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340251958413563e-05,
+      "loss": 4.1858,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339413363662511e-05,
+      "loss": 4.1998,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385764067918316e-05,
+      "loss": 4.1939,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377378120407796e-05,
+      "loss": 4.1986,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3368992172897276e-05,
+      "loss": 4.2003,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360606225386756e-05,
+      "loss": 4.194,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352236656679965e-05,
+      "loss": 4.2056,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3343850709169445e-05,
+      "loss": 4.1839,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333548114046266e-05,
+      "loss": 4.1764,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332709519295214e-05,
+      "loss": 4.1937,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331870924544162e-05,
+      "loss": 4.1881,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.1827,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330193735042058e-05,
+      "loss": 4.1991,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3293551402910054e-05,
+      "loss": 4.1801,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285165455399534e-05,
+      "loss": 4.1805,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3276779507889013e-05,
+      "loss": 4.1837,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326840993918223e-05,
+      "loss": 4.1785,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32600239916717e-05,
+      "loss": 4.194,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325163804416118e-05,
+      "loss": 4.1859,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324325209665067e-05,
+      "loss": 4.1802,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3234882527943885e-05,
+      "loss": 4.1813,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322649658043336e-05,
+      "loss": 4.1743,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321811063292284e-05,
+      "loss": 4.1943,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320972468541232e-05,
+      "loss": 4.1754,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320135511670553e-05,
+      "loss": 4.1919,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319296916919501e-05,
+      "loss": 4.1678,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318459960048822e-05,
+      "loss": 4.1858,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31762136529777e-05,
+      "loss": 4.1808,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167827705467176e-05,
+      "loss": 4.1777,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159441757956656e-05,
+      "loss": 4.1888,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151055810446136e-05,
+      "loss": 4.1831,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314266986293562e-05,
+      "loss": 4.1867,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31342839154251e-05,
+      "loss": 4.1968,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312589796791458e-05,
+      "loss": 4.1963,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311752839920779e-05,
+      "loss": 4.1864,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310914245169727e-05,
+      "loss": 4.179,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310075650418675e-05,
+      "loss": 4.1712,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309238693547996e-05,
+      "loss": 4.1829,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308400098796944e-05,
+      "loss": 4.1945,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307561504045892e-05,
+      "loss": 4.1876,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30672290929484e-05,
+      "loss": 4.1785,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305884314543788e-05,
+      "loss": 4.1645,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305045719792736e-05,
+      "loss": 4.1923,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304207125041684e-05,
+      "loss": 4.164,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303368530290632e-05,
+      "loss": 4.1781,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025315734199536e-05,
+      "loss": 4.179,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3016929786689016e-05,
+      "loss": 4.1849,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008543839178496e-05,
+      "loss": 4.182,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000157891667976e-05,
+      "loss": 4.1891,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991771944157456e-05,
+      "loss": 4.1662,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983402375450665e-05,
+      "loss": 4.176,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975016427940145e-05,
+      "loss": 4.1795,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966630480429625e-05,
+      "loss": 4.1702,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958244532919105e-05,
+      "loss": 4.1683,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949874964212314e-05,
+      "loss": 4.1882,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941489016701794e-05,
+      "loss": 4.1804,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933103069191274e-05,
+      "loss": 4.1881,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292473350048449e-05,
+      "loss": 4.1738,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291634755297397e-05,
+      "loss": 4.1793,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290796160546345e-05,
+      "loss": 4.1668,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289957565795293e-05,
+      "loss": 4.1819,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289118971044241e-05,
+      "loss": 4.1863,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288280376293189e-05,
+      "loss": 4.1687,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287441781542136e-05,
+      "loss": 4.1854,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286603186791084e-05,
+      "loss": 4.1524,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285767867800779e-05,
+      "loss": 4.1641,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284929273049727e-05,
+      "loss": 4.1634,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284090678298675e-05,
+      "loss": 4.1716,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283252083547623e-05,
+      "loss": 4.1799,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282413488796571e-05,
+      "loss": 4.174,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281574894045519e-05,
+      "loss": 4.1518,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280736299294467e-05,
+      "loss": 4.172,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279897704543415e-05,
+      "loss": 4.1792,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279059109792363e-05,
+      "loss": 4.1763,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278222152921684e-05,
+      "loss": 4.1758,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277385196051005e-05,
+      "loss": 4.1669,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276546601299953e-05,
+      "loss": 4.1665,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275708006548901e-05,
+      "loss": 4.1851,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748694117978486e-05,
+      "loss": 4.1688,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740308170467966e-05,
+      "loss": 4.1678,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2731922222957446e-05,
+      "loss": 4.1655,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272353627544693e-05,
+      "loss": 4.1616,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271515032793641e-05,
+      "loss": 4.168,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270678075922962e-05,
+      "loss": 4.1696,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26983948117191e-05,
+      "loss": 4.1772,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269000886420858e-05,
+      "loss": 4.1603,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268162291669806e-05,
+      "loss": 4.1621,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267323696918754e-05,
+      "loss": 4.1796,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2664883779284486e-05,
+      "loss": 4.164,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265649783177396e-05,
+      "loss": 4.1721,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264811188426344e-05,
+      "loss": 4.1741,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263972593675292e-05,
+      "loss": 4.18,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26313399892424e-05,
+      "loss": 4.1769,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2622954041731886e-05,
+      "loss": 4.1631,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2614568094221366e-05,
+      "loss": 4.1528,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606182146710846e-05,
+      "loss": 4.1623,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597812578004055e-05,
+      "loss": 4.1678,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589426630493535e-05,
+      "loss": 4.1768,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581040682983015e-05,
+      "loss": 4.1727,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2572654735472495e-05,
+      "loss": 4.1669,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564268787961975e-05,
+      "loss": 4.1677,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2555882840451455e-05,
+      "loss": 4.1657,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547496892940935e-05,
+      "loss": 4.1547,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539110945430415e-05,
+      "loss": 4.1621,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2530741376723624e-05,
+      "loss": 4.1743,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522355429213104e-05,
+      "loss": 4.1765,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251398586050632e-05,
+      "loss": 4.1576,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25055999129958e-05,
+      "loss": 4.157,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.159791946411133,
+      "eval_runtime": 563.6896,
+      "eval_samples_per_second": 676.952,
+      "eval_steps_per_second": 21.155,
+      "step": 457914
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249721396548528e-05,
+      "loss": 4.181,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248882801797476e-05,
+      "loss": 4.1661,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248044207046424e-05,
+      "loss": 4.1575,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247205612295372e-05,
+      "loss": 4.1581,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24636701754432e-05,
+      "loss": 4.1532,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245528422793267e-05,
+      "loss": 4.1517,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244689828042215e-05,
+      "loss": 4.151,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243851233291163e-05,
+      "loss": 4.1534,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243012638540111e-05,
+      "loss": 4.1593,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242174043789059e-05,
+      "loss": 4.1637,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241335449038007e-05,
+      "loss": 4.1632,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240496854286955e-05,
+      "loss": 4.1552,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239658259535903e-05,
+      "loss": 4.1687,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238819664784852e-05,
+      "loss": 4.159,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379810700338e-05,
+      "loss": 4.1586,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237142475282748e-05,
+      "loss": 4.1454,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236303880531696e-05,
+      "loss": 4.1471,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235465285780644e-05,
+      "loss": 4.1628,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234626691029591e-05,
+      "loss": 4.1573,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233788096278539e-05,
+      "loss": 4.167,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232951139407861e-05,
+      "loss": 4.1557,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232112544656809e-05,
+      "loss": 4.1632,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231273949905756e-05,
+      "loss": 4.1596,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230435355154704e-05,
+      "loss": 4.1576,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229598398284026e-05,
+      "loss": 4.153,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228759803532974e-05,
+      "loss": 4.1525,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227921208781922e-05,
+      "loss": 4.1604,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22708261403087e-05,
+      "loss": 4.1439,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226245657160191e-05,
+      "loss": 4.1621,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254070624091386e-05,
+      "loss": 4.1507,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245684676580866e-05,
+      "loss": 4.1433,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237298729070346e-05,
+      "loss": 4.1564,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2228912781559826e-05,
+      "loss": 4.1574,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220543212853035e-05,
+      "loss": 4.1612,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212157265342515e-05,
+      "loss": 4.148,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203771317831995e-05,
+      "loss": 4.1529,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195385370321475e-05,
+      "loss": 4.154,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218701580161469e-05,
+      "loss": 4.1642,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217862985410417e-05,
+      "loss": 4.1558,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217024390659365e-05,
+      "loss": 4.1406,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216187433788686e-05,
+      "loss": 4.1412,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215348839037634e-05,
+      "loss": 4.1392,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214510244286582e-05,
+      "loss": 4.1574,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21367164953553e-05,
+      "loss": 4.1492,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212833054784478e-05,
+      "loss": 4.148,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211994460033426e-05,
+      "loss": 4.1588,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211155865282374e-05,
+      "loss": 4.1459,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210317270531322e-05,
+      "loss": 4.1595,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209480313660643e-05,
+      "loss": 4.14,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208641718909591e-05,
+      "loss": 4.1324,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078031241585395e-05,
+      "loss": 4.147,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2069645294074875e-05,
+      "loss": 4.1419,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061259346564355e-05,
+      "loss": 4.1419,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052889777857564e-05,
+      "loss": 4.1492,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044503830347044e-05,
+      "loss": 4.1397,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203613426164025e-05,
+      "loss": 4.1365,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202774831412973e-05,
+      "loss": 4.137,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201936236661921e-05,
+      "loss": 4.1365,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201097641910869e-05,
+      "loss": 4.1477,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200259047159817e-05,
+      "loss": 4.1432,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199420452408765e-05,
+      "loss": 4.1338,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198581857657713e-05,
+      "loss": 4.1374,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197743262906661e-05,
+      "loss": 4.1259,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196904668155609e-05,
+      "loss": 4.1517,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196067711284931e-05,
+      "loss": 4.1347,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195229116533879e-05,
+      "loss": 4.1427,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1943921596632e-05,
+      "loss": 4.1304,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193553564912148e-05,
+      "loss": 4.1372,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192714970161096e-05,
+      "loss": 4.1386,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191876375410044e-05,
+      "loss": 4.1348,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191037780658992e-05,
+      "loss": 4.1467,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19019918590794e-05,
+      "loss": 4.1374,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189360591156887e-05,
+      "loss": 4.1449,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188521996405835e-05,
+      "loss": 4.1535,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187683401654783e-05,
+      "loss": 4.1497,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186844806903731e-05,
+      "loss": 4.1462,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860078500330526e-05,
+      "loss": 4.1362,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185170893162374e-05,
+      "loss": 4.1291,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184332298411322e-05,
+      "loss": 4.1379,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1834937036602695e-05,
+      "loss": 4.151,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826551089092175e-05,
+      "loss": 4.1448,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818165141581655e-05,
+      "loss": 4.1352,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809779194071135e-05,
+      "loss": 4.1216,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801393246560615e-05,
+      "loss": 4.15,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793023677853824e-05,
+      "loss": 4.1189,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784637730343304e-05,
+      "loss": 4.1362,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776251782832784e-05,
+      "loss": 4.1385,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767865835322264e-05,
+      "loss": 4.1371,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175947988781175e-05,
+      "loss": 4.1422,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175109394030123e-05,
+      "loss": 4.1464,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174270799279071e-05,
+      "loss": 4.1213,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173433842408392e-05,
+      "loss": 4.1357,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17259524765734e-05,
+      "loss": 4.1376,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171756652906288e-05,
+      "loss": 4.1276,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170918058155236e-05,
+      "loss": 4.1228,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170079463404184e-05,
+      "loss": 4.1463,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169240868653132e-05,
+      "loss": 4.139,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16840227390208e-05,
+      "loss": 4.148,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167563679151028e-05,
+      "loss": 4.1301,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166726722280349e-05,
+      "loss": 4.138,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165888127529297e-05,
+      "loss": 4.1233,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165049532778245e-05,
+      "loss": 4.1389,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642109380271935e-05,
+      "loss": 4.1444,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633739811565144e-05,
+      "loss": 4.1327,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625353864054624e-05,
+      "loss": 4.1416,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1616967916544104e-05,
+      "loss": 4.1144,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1608581969033584e-05,
+      "loss": 4.119,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160019602152306e-05,
+      "loss": 4.1192,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159182645281627e-05,
+      "loss": 4.1307,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158344050530575e-05,
+      "loss": 4.1379,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575054557795226e-05,
+      "loss": 4.1362,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1566668610284706e-05,
+      "loss": 4.1105,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155829904157792e-05,
+      "loss": 4.129,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15499130940674e-05,
+      "loss": 4.1379,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154152714655688e-05,
+      "loss": 4.138,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153314119904636e-05,
+      "loss": 4.1346,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152475525153584e-05,
+      "loss": 4.1262,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151636930402532e-05,
+      "loss": 4.1291,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15079833565148e-05,
+      "loss": 4.1375,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149961378780801e-05,
+      "loss": 4.1308,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149122784029749e-05,
+      "loss": 4.1271,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148284189278697e-05,
+      "loss": 4.1242,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147447232408018e-05,
+      "loss": 4.1211,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146608637656966e-05,
+      "loss": 4.1299,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145770042905914e-05,
+      "loss": 4.1304,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449330860352356e-05,
+      "loss": 4.1332,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1440944912841836e-05,
+      "loss": 4.1226,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432558965331316e-05,
+      "loss": 4.1207,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424173017820796e-05,
+      "loss": 4.1387,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415787070310276e-05,
+      "loss": 4.121,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407401122799756e-05,
+      "loss": 4.1338,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399015175289236e-05,
+      "loss": 4.1363,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390629227778716e-05,
+      "loss": 4.1356,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382243280268195e-05,
+      "loss": 4.1387,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1373857332757675e-05,
+      "loss": 4.1283,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1365471385247155e-05,
+      "loss": 4.1128,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357085437736635e-05,
+      "loss": 4.1214,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1348699490226115e-05,
+      "loss": 4.1253,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340313542715595e-05,
+      "loss": 4.1399,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1331927595205075e-05,
+      "loss": 4.1311,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323541647694555e-05,
+      "loss": 4.1326,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13151884577915e-05,
+      "loss": 4.1244,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130680251028098e-05,
+      "loss": 4.1265,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129841656277046e-05,
+      "loss": 4.1138,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129003061525994e-05,
+      "loss": 4.1251,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128164466774942e-05,
+      "loss": 4.1351,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127325872023889e-05,
+      "loss": 4.1385,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126487277272837e-05,
+      "loss": 4.1167,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125650320402159e-05,
+      "loss": 4.1211,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.128833770751953,
+      "eval_runtime": 555.1656,
+      "eval_samples_per_second": 687.346,
+      "eval_steps_per_second": 21.48,
+      "step": 534233
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124811725651106e-05,
+      "loss": 4.145,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123973130900054e-05,
+      "loss": 4.1264,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123134536149003e-05,
+      "loss": 4.1197,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122295941397951e-05,
+      "loss": 4.1226,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121457346646899e-05,
+      "loss": 4.1107,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120618751895847e-05,
+      "loss": 4.1166,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119780157144795e-05,
+      "loss": 4.108,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118941562393743e-05,
+      "loss": 4.1179,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118102967642691e-05,
+      "loss": 4.1178,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117264372891639e-05,
+      "loss": 4.1262,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116425778140587e-05,
+      "loss": 4.1232,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115587183389535e-05,
+      "loss": 4.121,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114748588638483e-05,
+      "loss": 4.1314,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11390999388743e-05,
+      "loss": 4.1163,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113071399136378e-05,
+      "loss": 4.1266,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112232804385326e-05,
+      "loss": 4.1036,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111395847514648e-05,
+      "loss": 4.1105,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110557252763596e-05,
+      "loss": 4.1231,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109718658012544e-05,
+      "loss": 4.1169,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108880063261492e-05,
+      "loss": 4.1296,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10804146851044e-05,
+      "loss": 4.1186,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107202873759388e-05,
+      "loss": 4.1244,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106364279008336e-05,
+      "loss": 4.1221,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105525684257284e-05,
+      "loss": 4.1172,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104687089506232e-05,
+      "loss": 4.121,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1038501326355526e-05,
+      "loss": 4.111,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030115378845006e-05,
+      "loss": 4.1236,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1021729431334486e-05,
+      "loss": 4.1142,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1013343483823966e-05,
+      "loss": 4.1225,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100497391511718e-05,
+      "loss": 4.1103,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099658796760666e-05,
+      "loss": 4.1089,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098820202009614e-05,
+      "loss": 4.119,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097981607258562e-05,
+      "loss": 4.1174,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09714301250751e-05,
+      "loss": 4.121,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096306055636831e-05,
+      "loss": 4.117,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095467460885779e-05,
+      "loss": 4.1112,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946305040151e-05,
+      "loss": 4.1186,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093791909264048e-05,
+      "loss": 4.1267,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092953314512996e-05,
+      "loss": 4.1183,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092114719761944e-05,
+      "loss": 4.107,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091277762891265e-05,
+      "loss": 4.1029,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0904391681402136e-05,
+      "loss": 4.1029,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896005733891615e-05,
+      "loss": 4.1205,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0887619786381095e-05,
+      "loss": 4.1135,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879233838870575e-05,
+      "loss": 4.1083,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0870864270163784e-05,
+      "loss": 4.1255,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0862478322653264e-05,
+      "loss": 4.1083,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0854092375142744e-05,
+      "loss": 4.1238,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0845706427632224e-05,
+      "loss": 4.1054,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0837320480121704e-05,
+      "loss": 4.0974,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0828934532611184e-05,
+      "loss": 4.11,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082056496390439e-05,
+      "loss": 4.111,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081217901639387e-05,
+      "loss": 4.1011,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080379306888335e-05,
+      "loss": 4.1163,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079540712137283e-05,
+      "loss": 4.1064,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078702117386231e-05,
+      "loss": 4.0981,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077863522635179e-05,
+      "loss": 4.1069,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077024927884127e-05,
+      "loss": 4.0944,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076186333133075e-05,
+      "loss": 4.1115,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075349376262397e-05,
+      "loss": 4.1069,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074512419391718e-05,
+      "loss": 4.1011,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073673824640666e-05,
+      "loss": 4.1006,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072835229889614e-05,
+      "loss": 4.0924,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071996635138561e-05,
+      "loss": 4.116,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071158040387509e-05,
+      "loss": 4.099,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070319445636457e-05,
+      "loss": 4.1088,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069480850885405e-05,
+      "loss": 4.0972,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068642256134354e-05,
+      "loss": 4.1027,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067805299263675e-05,
+      "loss": 4.0994,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066966704512623e-05,
+      "loss": 4.1032,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066128109761571e-05,
+      "loss": 4.1087,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065289515010519e-05,
+      "loss": 4.1034,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064450920259467e-05,
+      "loss": 4.1085,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636139633887876e-05,
+      "loss": 4.1177,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0627753686377356e-05,
+      "loss": 4.1167,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0619367738866836e-05,
+      "loss": 4.1067,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0610981791356316e-05,
+      "loss": 4.1072,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0602595843845796e-05,
+      "loss": 4.0884,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059424265394274e-05,
+      "loss": 4.1057,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058585670643222e-05,
+      "loss": 4.1152,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05774707589217e-05,
+      "loss": 4.1096,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056908481141118e-05,
+      "loss": 4.1053,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056069886390066e-05,
+      "loss": 4.0869,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055231291639014e-05,
+      "loss": 4.1105,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054392696887962e-05,
+      "loss": 4.0893,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053555740017283e-05,
+      "loss": 4.0999,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052717145266231e-05,
+      "loss": 4.1007,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051880188395552e-05,
+      "loss": 4.1043,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0510415936445e-05,
+      "loss": 4.1092,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050202998893448e-05,
+      "loss": 4.1082,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049364404142396e-05,
+      "loss": 4.0912,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0485274472717174e-05,
+      "loss": 4.1048,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0476888525206654e-05,
+      "loss": 4.0992,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0468502577696134e-05,
+      "loss": 4.0951,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460116630185614e-05,
+      "loss": 4.089,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0451730682675094e-05,
+      "loss": 4.1101,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443344735164574e-05,
+      "loss": 4.1055,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0434958787654054e-05,
+      "loss": 4.1133,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0426572840143534e-05,
+      "loss": 4.0992,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418186892633014e-05,
+      "loss": 4.1044,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0409800945122494e-05,
+      "loss": 4.091,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0401414997611974e-05,
+      "loss": 4.1078,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039302905010145e-05,
+      "loss": 4.1064,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038464310259093e-05,
+      "loss": 4.1017,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0376257155080414e-05,
+      "loss": 4.104,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0367871207569894e-05,
+      "loss": 4.0854,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0359485260059374e-05,
+      "loss": 4.0844,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035111569135258e-05,
+      "loss": 4.0887,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034272974384206e-05,
+      "loss": 4.0948,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033436017513527e-05,
+      "loss": 4.108,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032597422762475e-05,
+      "loss": 4.1004,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031758828011423e-05,
+      "loss": 4.0738,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030920233260371e-05,
+      "loss": 4.0996,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030081638509319e-05,
+      "loss": 4.1048,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029243043758267e-05,
+      "loss": 4.1035,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028406086887588e-05,
+      "loss": 4.1016,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027567492136537e-05,
+      "loss": 4.0962,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026728897385485e-05,
+      "loss": 4.0904,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025890302634433e-05,
+      "loss": 4.1058,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0250533457637536e-05,
+      "loss": 4.0997,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242147510127016e-05,
+      "loss": 4.0949,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0233761562616496e-05,
+      "loss": 4.0897,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0225375615105976e-05,
+      "loss": 4.0905,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0216989667595456e-05,
+      "loss": 4.0961,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0208603720084936e-05,
+      "loss": 4.0946,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0200234151378145e-05,
+      "loss": 4.1055,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0191848203867625e-05,
+      "loss": 4.0911,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0183462256357105e-05,
+      "loss": 4.089,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175076308846585e-05,
+      "loss": 4.1077,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0166690361336065e-05,
+      "loss": 4.0877,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015832079262928e-05,
+      "loss": 4.1028,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014993484511876e-05,
+      "loss": 4.1041,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014154889760824e-05,
+      "loss": 4.1005,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013316295009772e-05,
+      "loss": 4.1062,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01247770025872e-05,
+      "loss": 4.0957,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011639105507668e-05,
+      "loss": 4.0828,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010800510756616e-05,
+      "loss": 4.0884,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0099619160055634e-05,
+      "loss": 4.0971,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0091233212545114e-05,
+      "loss": 4.1038,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0082847265034594e-05,
+      "loss": 4.1015,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0074461317524074e-05,
+      "loss": 4.0986,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006609174881728e-05,
+      "loss": 4.0909,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005770580130677e-05,
+      "loss": 4.0939,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004931985379625e-05,
+      "loss": 4.0827,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004093390628573e-05,
+      "loss": 4.0935,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003256433757894e-05,
+      "loss": 4.1021,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002417839006842e-05,
+      "loss": 4.1075,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00157924425579e-05,
+      "loss": 4.0823,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000740649504738e-05,
+      "loss": 4.0934,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.104266166687012,
+      "eval_runtime": 541.9624,
+      "eval_samples_per_second": 704.091,
+      "eval_steps_per_second": 22.003,
+      "step": 610552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999902054753686e-05,
+      "loss": 4.1015,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999065097883007e-05,
+      "loss": 4.0983,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998226503131955e-05,
+      "loss": 4.0896,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997387908380903e-05,
+      "loss": 4.0903,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996549313629851e-05,
+      "loss": 4.0786,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995712356759172e-05,
+      "loss": 4.083,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99487376200812e-05,
+      "loss": 4.0814,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994035167257068e-05,
+      "loss": 4.0822,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.993196572506016e-05,
+      "loss": 4.0853,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992357977754964e-05,
+      "loss": 4.095,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991522658764658e-05,
+      "loss": 4.0895,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.990684064013606e-05,
+      "loss": 4.0914,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989845469262554e-05,
+      "loss": 4.1012,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989006874511502e-05,
+      "loss": 4.0816,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98816827976045e-05,
+      "loss": 4.0989,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987329685009398e-05,
+      "loss": 4.075,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986492728138719e-05,
+      "loss": 4.0786,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.985654133387668e-05,
+      "loss": 4.0917,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984815538636616e-05,
+      "loss": 4.0872,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983976943885564e-05,
+      "loss": 4.0976,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983138349134512e-05,
+      "loss": 4.0888,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9822997543834597e-05,
+      "loss": 4.0899,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9814611596324077e-05,
+      "loss": 4.0922,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9806225648813557e-05,
+      "loss": 4.086,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9797856080106766e-05,
+      "loss": 4.0915,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9789486511399975e-05,
+      "loss": 4.0804,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781100563889455e-05,
+      "loss": 4.0926,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9772714616378935e-05,
+      "loss": 4.0805,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9764328668868415e-05,
+      "loss": 4.0949,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9755942721357895e-05,
+      "loss": 4.0772,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9747556773847374e-05,
+      "loss": 4.0789,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973917082633686e-05,
+      "loss": 4.0894,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973078487882634e-05,
+      "loss": 4.0874,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.972239893131582e-05,
+      "loss": 4.0912,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9714012983805294e-05,
+      "loss": 4.0835,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970564341509851e-05,
+      "loss": 4.0858,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969727384639172e-05,
+      "loss": 4.0856,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96888878988812e-05,
+      "loss": 4.0949,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968050195137068e-05,
+      "loss": 4.0912,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967211600386016e-05,
+      "loss": 4.0773,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966373005634964e-05,
+      "loss": 4.0734,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965534410883912e-05,
+      "loss": 4.0727,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96469581613286e-05,
+      "loss": 4.0888,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963857221381808e-05,
+      "loss": 4.0879,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9630202645111295e-05,
+      "loss": 4.0732,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962181669760077e-05,
+      "loss": 4.0972,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961343075009025e-05,
+      "loss": 4.077,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960504480257973e-05,
+      "loss": 4.0934,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9596675233872944e-05,
+      "loss": 4.0768,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958828928636242e-05,
+      "loss": 4.0714,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95799033388519e-05,
+      "loss": 4.0776,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957151739134138e-05,
+      "loss": 4.0777,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956313144383086e-05,
+      "loss": 4.0802,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955474549632034e-05,
+      "loss": 4.0808,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9546375927613546e-05,
+      "loss": 4.0795,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953798998010303e-05,
+      "loss": 4.0669,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952960403259251e-05,
+      "loss": 4.0764,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952123446388572e-05,
+      "loss": 4.068,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95128485163752e-05,
+      "loss": 4.0807,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950446256886468e-05,
+      "loss": 4.072,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949607662135416e-05,
+      "loss": 4.0761,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948769067384364e-05,
+      "loss": 4.0739,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947930472633312e-05,
+      "loss": 4.0641,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947093515762633e-05,
+      "loss": 4.0845,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946254921011581e-05,
+      "loss": 4.0698,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945416326260529e-05,
+      "loss": 4.081,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944577731509477e-05,
+      "loss": 4.0695,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943739136758425e-05,
+      "loss": 4.072,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942900542007373e-05,
+      "loss": 4.0697,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942061947256322e-05,
+      "loss": 4.0727,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94122335250527e-05,
+      "loss": 4.0786,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940384757754218e-05,
+      "loss": 4.0766,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9395478008835386e-05,
+      "loss": 4.0774,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9387092061324866e-05,
+      "loss": 4.0955,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9378706113814346e-05,
+      "loss": 4.085,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9370320166303826e-05,
+      "loss": 4.0766,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9361950597597035e-05,
+      "loss": 4.0789,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9353564650086515e-05,
+      "loss": 4.0615,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9345195081379724e-05,
+      "loss": 4.0732,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9336809133869204e-05,
+      "loss": 4.0867,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9328423186358684e-05,
+      "loss": 4.0855,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932003723884817e-05,
+      "loss": 4.0745,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931165129133765e-05,
+      "loss": 4.0599,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930326534382713e-05,
+      "loss": 4.0805,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9294879396316604e-05,
+      "loss": 4.0574,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928650982760982e-05,
+      "loss": 4.0738,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92781238800993e-05,
+      "loss": 4.0735,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926973793258878e-05,
+      "loss": 4.0711,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926135198507825e-05,
+      "loss": 4.0824,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925298241637147e-05,
+      "loss": 4.0811,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924459646886095e-05,
+      "loss": 4.065,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923621052135042e-05,
+      "loss": 4.0698,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922782457383991e-05,
+      "loss": 4.0736,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921943862632939e-05,
+      "loss": 4.067,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921105267881887e-05,
+      "loss": 4.0643,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920266673130835e-05,
+      "loss": 4.0763,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919428078379783e-05,
+      "loss": 4.0782,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918591121509104e-05,
+      "loss": 4.0887,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917752526758052e-05,
+      "loss": 4.0704,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916913932007e-05,
+      "loss": 4.0773,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916075337255948e-05,
+      "loss": 4.0648,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9152383803852686e-05,
+      "loss": 4.0741,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9143997856342166e-05,
+      "loss": 4.083,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9135611908831646e-05,
+      "loss": 4.0744,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9127225961321126e-05,
+      "loss": 4.0785,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911885639261434e-05,
+      "loss": 4.054,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911047044510382e-05,
+      "loss": 4.0578,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91020844975933e-05,
+      "loss": 4.0612,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909369855008278e-05,
+      "loss": 4.07,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908532898137599e-05,
+      "loss": 4.0727,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.907694303386547e-05,
+      "loss": 4.0758,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906855708635495e-05,
+      "loss": 4.0454,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906017113884443e-05,
+      "loss": 4.0727,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905178519133391e-05,
+      "loss": 4.0804,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904339924382339e-05,
+      "loss": 4.0724,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903501329631287e-05,
+      "loss": 4.0739,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902662734880235e-05,
+      "loss": 4.073,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901824140129183e-05,
+      "loss": 4.0602,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900987183258504e-05,
+      "loss": 4.0752,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9001518642681985e-05,
+      "loss": 4.0785,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8993132695171465e-05,
+      "loss": 4.0615,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8984746747660945e-05,
+      "loss": 4.0631,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8976360800150425e-05,
+      "loss": 4.0619,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8967974852639905e-05,
+      "loss": 4.0666,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8959588905129385e-05,
+      "loss": 4.0665,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8951202957618865e-05,
+      "loss": 4.0738,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8942817010108345e-05,
+      "loss": 4.0686,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8934431062597825e-05,
+      "loss": 4.0584,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8926045115087304e-05,
+      "loss": 4.0811,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8917675546380514e-05,
+      "loss": 4.059,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8909289598869994e-05,
+      "loss": 4.0753,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890090365135948e-05,
+      "loss": 4.0793,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889251770384896e-05,
+      "loss": 4.0732,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.888413175633844e-05,
+      "loss": 4.0785,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887574580882791e-05,
+      "loss": 4.0692,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886735986131739e-05,
+      "loss": 4.0552,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885897391380687e-05,
+      "loss": 4.0631,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885058796629635e-05,
+      "loss": 4.0652,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884221839758956e-05,
+      "loss": 4.0783,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883383245007904e-05,
+      "loss": 4.0737,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882544650256852e-05,
+      "loss": 4.0701,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881707693386173e-05,
+      "loss": 4.0699,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880869098635122e-05,
+      "loss": 4.0646,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88003050388407e-05,
+      "loss": 4.0541,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879191909133018e-05,
+      "loss": 4.0671,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878353314381966e-05,
+      "loss": 4.0756,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877514719630914e-05,
+      "loss": 4.0816,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876676124879862e-05,
+      "loss": 4.0594,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.875839168009183e-05,
+      "loss": 4.0625,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.08506965637207,
+      "eval_runtime": 541.3196,
+      "eval_samples_per_second": 704.927,
+      "eval_steps_per_second": 22.029,
+      "step": 686871
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.875000573258131e-05,
+      "loss": 4.0861,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874161978507079e-05,
+      "loss": 4.0696,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873323383756027e-05,
+      "loss": 4.0655,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872484789004975e-05,
+      "loss": 4.062,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871646194253923e-05,
+      "loss": 4.0561,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870807599502871e-05,
+      "loss": 4.056,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869969004751819e-05,
+      "loss": 4.0549,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8691304100007667e-05,
+      "loss": 4.0561,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8682918152497147e-05,
+      "loss": 4.0568,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8674532204986626e-05,
+      "loss": 4.0702,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866616263627984e-05,
+      "loss": 4.0618,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865777668876932e-05,
+      "loss": 4.0643,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86493907412588e-05,
+      "loss": 4.0748,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8641004793748275e-05,
+      "loss": 4.0586,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8632618846237755e-05,
+      "loss": 4.0692,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8624232898727235e-05,
+      "loss": 4.0488,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8615846951216715e-05,
+      "loss": 4.0523,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8607461003706195e-05,
+      "loss": 4.0682,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599075056195675e-05,
+      "loss": 4.0604,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8590689108685155e-05,
+      "loss": 4.0738,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8582303161174635e-05,
+      "loss": 4.0564,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8573917213664115e-05,
+      "loss": 4.0744,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8565531266153595e-05,
+      "loss": 4.0589,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855714531864308e-05,
+      "loss": 4.0603,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854875937113256e-05,
+      "loss": 4.0642,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854037342362204e-05,
+      "loss": 4.0602,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8531987476111515e-05,
+      "loss": 4.0662,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852361790740473e-05,
+      "loss": 4.054,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851523195989421e-05,
+      "loss": 4.0696,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850684601238369e-05,
+      "loss": 4.0555,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8498460064873164e-05,
+      "loss": 4.047,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849009049616638e-05,
+      "loss": 4.0655,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.848170454865586e-05,
+      "loss": 4.0607,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.847333497994907e-05,
+      "loss": 4.0659,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846494903243855e-05,
+      "loss": 4.0602,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8456563084928035e-05,
+      "loss": 4.0557,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8448177137417515e-05,
+      "loss": 4.0621,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843979118990699e-05,
+      "loss": 4.066,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843140524239647e-05,
+      "loss": 4.0686,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842301929488595e-05,
+      "loss": 4.0469,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841463334737543e-05,
+      "loss": 4.0524,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840626377866864e-05,
+      "loss": 4.0495,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839787783115812e-05,
+      "loss": 4.0571,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83894918836476e-05,
+      "loss": 4.0634,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838110593613708e-05,
+      "loss": 4.0475,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837271998862656e-05,
+      "loss": 4.0696,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836435041991977e-05,
+      "loss": 4.0506,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835596447240925e-05,
+      "loss": 4.0687,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834759490370246e-05,
+      "loss": 4.0538,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833920895619194e-05,
+      "loss": 4.0461,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833082300868142e-05,
+      "loss": 4.0557,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83224370611709e-05,
+      "loss": 4.0452,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831405111366038e-05,
+      "loss": 4.0579,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830568154495359e-05,
+      "loss": 4.0526,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829729559744307e-05,
+      "loss": 4.0564,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828890964993255e-05,
+      "loss": 4.0381,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828052370242203e-05,
+      "loss": 4.0524,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827213775491151e-05,
+      "loss": 4.0386,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826375180740099e-05,
+      "loss": 4.058,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825536585989047e-05,
+      "loss": 4.0482,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824697991237996e-05,
+      "loss": 4.05,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823861034367317e-05,
+      "loss": 4.0486,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823022439616265e-05,
+      "loss": 4.0391,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.822183844865213e-05,
+      "loss": 4.0576,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821345250114161e-05,
+      "loss": 4.0484,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8205066553631087e-05,
+      "loss": 4.0584,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8196680606120567e-05,
+      "loss": 4.0416,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8188294658610047e-05,
+      "loss": 4.0489,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8179925089903256e-05,
+      "loss": 4.0434,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171539142392736e-05,
+      "loss": 4.0519,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8163169573685945e-05,
+      "loss": 4.0507,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154783626175425e-05,
+      "loss": 4.0524,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814639767866491e-05,
+      "loss": 4.054,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813801173115439e-05,
+      "loss": 4.072,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812962578364387e-05,
+      "loss": 4.0565,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812123983613335e-05,
+      "loss": 4.0541,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8112853888622824e-05,
+      "loss": 4.0525,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810448431991604e-05,
+      "loss": 4.0415,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809609837240552e-05,
+      "loss": 4.0451,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8087712424894993e-05,
+      "loss": 4.0585,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807932647738447e-05,
+      "loss": 4.0641,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807094052987395e-05,
+      "loss": 4.0545,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806255458236343e-05,
+      "loss": 4.0355,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805416863485291e-05,
+      "loss": 4.0506,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804578268734239e-05,
+      "loss": 4.0377,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803741311863561e-05,
+      "loss": 4.0465,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802902717112509e-05,
+      "loss": 4.0489,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802064122361457e-05,
+      "loss": 4.0482,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801225527610405e-05,
+      "loss": 4.0573,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800386932859353e-05,
+      "loss": 4.0536,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799548338108301e-05,
+      "loss": 4.0448,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798711381237622e-05,
+      "loss": 4.0467,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79787278648657e-05,
+      "loss": 4.0524,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797034191735518e-05,
+      "loss": 4.0389,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796195596984466e-05,
+      "loss": 4.043,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795357002233414e-05,
+      "loss": 4.0489,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794520045362735e-05,
+      "loss": 4.0557,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793681450611683e-05,
+      "loss": 4.0625,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7928428558606314e-05,
+      "loss": 4.0441,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920042611095793e-05,
+      "loss": 4.0541,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911673042389e-05,
+      "loss": 4.043,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.790328709487848e-05,
+      "loss": 4.053,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789490114736796e-05,
+      "loss": 4.0513,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788651519985744e-05,
+      "loss": 4.0505,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787812925234692e-05,
+      "loss": 4.0539,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78697433048364e-05,
+      "loss": 4.032,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786135735732588e-05,
+      "loss": 4.0359,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7852971409815355e-05,
+      "loss": 4.0373,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784460184110857e-05,
+      "loss": 4.0429,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783621589359805e-05,
+      "loss": 4.0488,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782782994608753e-05,
+      "loss": 4.0505,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781944399857701e-05,
+      "loss": 4.0285,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781105805106649e-05,
+      "loss": 4.0473,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780267210355597e-05,
+      "loss": 4.0547,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779430253484919e-05,
+      "loss": 4.0487,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778591658733866e-05,
+      "loss": 4.0517,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777753063982814e-05,
+      "loss": 4.0457,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776914469231762e-05,
+      "loss": 4.0394,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.77607587448071e-05,
+      "loss": 4.0482,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775238917610031e-05,
+      "loss": 4.0553,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774400322858979e-05,
+      "loss": 4.0439,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7735633659883005e-05,
+      "loss": 4.0356,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7727247712372485e-05,
+      "loss": 4.0422,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7718861764861965e-05,
+      "loss": 4.0412,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771049219615518e-05,
+      "loss": 4.0462,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770210624864466e-05,
+      "loss": 4.043,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7693720301134134e-05,
+      "loss": 4.0504,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7685334353623614e-05,
+      "loss": 4.0355,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7676948406113094e-05,
+      "loss": 4.0566,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7668562458602574e-05,
+      "loss": 4.0378,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7660176511092054e-05,
+      "loss": 4.0499,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7651790563581534e-05,
+      "loss": 4.0558,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7643404616071014e-05,
+      "loss": 4.049,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763503504736422e-05,
+      "loss": 4.0553,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76266490998537e-05,
+      "loss": 4.048,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761826315234319e-05,
+      "loss": 4.0315,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760987720483267e-05,
+      "loss": 4.0398,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760149125732215e-05,
+      "loss": 4.0365,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759310530981163e-05,
+      "loss": 4.056,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758471936230111e-05,
+      "loss": 4.0524,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757633341479059e-05,
+      "loss": 4.0472,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756794746728007e-05,
+      "loss": 4.0479,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755956151976954e-05,
+      "loss": 4.0394,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755119195106276e-05,
+      "loss": 4.0299,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754280600355224e-05,
+      "loss": 4.0431,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753442005604172e-05,
+      "loss": 4.0548,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752603410853119e-05,
+      "loss": 4.0548,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751764816102067e-05,
+      "loss": 4.0365,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.750927859231389e-05,
+      "loss": 4.0403,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.069027423858643,
+      "eval_runtime": 557.3678,
+      "eval_samples_per_second": 684.631,
+      "eval_steps_per_second": 21.395,
+      "step": 763190
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750089264480337e-05,
+      "loss": 4.0757,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749250669729285e-05,
+      "loss": 4.0413,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748412074978233e-05,
+      "loss": 4.0471,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747573480227181e-05,
+      "loss": 4.0363,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746734885476129e-05,
+      "loss": 4.0413,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745896290725077e-05,
+      "loss": 4.0302,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745057695974025e-05,
+      "loss": 4.0358,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744219101222973e-05,
+      "loss": 4.0284,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743380506471921e-05,
+      "loss": 4.0356,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742541911720869e-05,
+      "loss": 4.0435,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417049548501896e-05,
+      "loss": 4.0428,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7408663600991376e-05,
+      "loss": 4.0419,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7400277653480856e-05,
+      "loss": 4.0495,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7391891705970336e-05,
+      "loss": 4.036,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738350575845982e-05,
+      "loss": 4.0485,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73751198109493e-05,
+      "loss": 4.0256,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736673386343878e-05,
+      "loss": 4.0266,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735834791592826e-05,
+      "loss": 4.0475,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7349961968417735e-05,
+      "loss": 4.0385,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7341576020907215e-05,
+      "loss": 4.0505,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7333190073396695e-05,
+      "loss": 4.0318,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7324804125886175e-05,
+      "loss": 4.0533,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7316418178375655e-05,
+      "loss": 4.0387,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7308032230865135e-05,
+      "loss": 4.0387,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7299662662158344e-05,
+      "loss": 4.041,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7291276714647824e-05,
+      "loss": 4.0388,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7282890767137304e-05,
+      "loss": 4.0438,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727450481962679e-05,
+      "loss": 4.03,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726611887211627e-05,
+      "loss": 4.0455,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725773292460575e-05,
+      "loss": 4.0384,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724934697709523e-05,
+      "loss": 4.0238,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724096102958471e-05,
+      "loss": 4.0407,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723257508207419e-05,
+      "loss": 4.039,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722422189217113e-05,
+      "loss": 4.0427,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721583594466061e-05,
+      "loss": 4.0361,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720744999715009e-05,
+      "loss": 4.0383,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719906404963957e-05,
+      "loss": 4.0405,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719069448093278e-05,
+      "loss": 4.0419,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718230853342226e-05,
+      "loss": 4.048,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7173922585911745e-05,
+      "loss": 4.0261,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7165536638401225e-05,
+      "loss": 4.0304,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157150690890705e-05,
+      "loss": 4.0279,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7148764743380184e-05,
+      "loss": 4.0363,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140378795869664e-05,
+      "loss": 4.0394,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7131992848359144e-05,
+      "loss": 4.0269,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7123623279652353e-05,
+      "loss": 4.0463,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7115237332141833e-05,
+      "loss": 4.0311,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7106851384631313e-05,
+      "loss": 4.0464,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709846543712079e-05,
+      "loss": 4.0362,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7090079489610267e-05,
+      "loss": 4.02,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7081693542099746e-05,
+      "loss": 4.032,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7073307594589226e-05,
+      "loss": 4.0247,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7064921647078706e-05,
+      "loss": 4.0393,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705655207837192e-05,
+      "loss": 4.0314,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70481661308614e-05,
+      "loss": 4.0331,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703978018335088e-05,
+      "loss": 4.0152,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703139423584036e-05,
+      "loss": 4.0341,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702300828832984e-05,
+      "loss": 4.0204,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701463871962305e-05,
+      "loss": 4.0383,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700626915091627e-05,
+      "loss": 4.0274,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699788320340574e-05,
+      "loss": 4.0282,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698949725589522e-05,
+      "loss": 4.0287,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69811113083847e-05,
+      "loss": 4.0152,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697272536087418e-05,
+      "loss": 4.0344,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696433941336366e-05,
+      "loss": 4.0273,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695595346585315e-05,
+      "loss": 4.0376,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694756751834263e-05,
+      "loss": 4.022,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6939197949635836e-05,
+      "loss": 4.0248,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6930812002125316e-05,
+      "loss": 4.0228,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922426054614796e-05,
+      "loss": 4.0289,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6914040107104276e-05,
+      "loss": 4.0276,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6905654159593756e-05,
+      "loss": 4.0331,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6897284590886965e-05,
+      "loss": 4.0276,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6888898643376445e-05,
+      "loss": 4.0497,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6880512695865925e-05,
+      "loss": 4.0386,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6872126748355405e-05,
+      "loss": 4.0313,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6863757179648614e-05,
+      "loss": 4.0342,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68553712321381e-05,
+      "loss": 4.0184,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684698528462758e-05,
+      "loss": 4.0296,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683859933711706e-05,
+      "loss": 4.0292,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683021338960654e-05,
+      "loss": 4.0413,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682182744209602e-05,
+      "loss": 4.034,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68134414945855e-05,
+      "loss": 4.0175,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680505554707498e-05,
+      "loss": 4.0277,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679668597836819e-05,
+      "loss": 4.0171,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.67883164096614e-05,
+      "loss": 4.0257,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677993046215088e-05,
+      "loss": 4.0249,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677154451464036e-05,
+      "loss": 4.0305,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676315856712984e-05,
+      "loss": 4.0379,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675477261961932e-05,
+      "loss": 4.0332,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67463866721088e-05,
+      "loss": 4.0192,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738000724598285e-05,
+      "loss": 4.0277,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729631155891494e-05,
+      "loss": 4.0304,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6721245208380974e-05,
+      "loss": 4.0179,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6712859260870454e-05,
+      "loss": 4.0191,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670447331335993e-05,
+      "loss": 4.0301,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669608736584941e-05,
+      "loss": 4.0381,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668770141833889e-05,
+      "loss": 4.0398,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667931547082837e-05,
+      "loss": 4.0253,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6670945902121576e-05,
+      "loss": 4.0322,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662559954611056e-05,
+      "loss": 4.0242,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6654174007100536e-05,
+      "loss": 4.0308,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664578805959002e-05,
+      "loss": 4.0285,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66374021120795e-05,
+      "loss": 4.0367,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662903254337271e-05,
+      "loss": 4.0305,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662064659586219e-05,
+      "loss": 4.0093,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661226064835167e-05,
+      "loss": 4.0163,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660387470084115e-05,
+      "loss": 4.0177,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659550513213436e-05,
+      "loss": 4.0236,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658711918462384e-05,
+      "loss": 4.0274,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657873323711332e-05,
+      "loss": 4.033,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65703472896028e-05,
+      "loss": 4.0092,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656196134209228e-05,
+      "loss": 4.022,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655357539458176e-05,
+      "loss": 4.0335,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654518944707124e-05,
+      "loss": 4.0284,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653680349956072e-05,
+      "loss": 4.0317,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6528433930853936e-05,
+      "loss": 4.0261,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6520047983343416e-05,
+      "loss": 4.0132,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6511662035832896e-05,
+      "loss": 4.027,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6503292467126105e-05,
+      "loss": 4.0385,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6494906519615585e-05,
+      "loss": 4.0212,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6486520572105065e-05,
+      "loss": 4.0148,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6478134624594545e-05,
+      "loss": 4.0212,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6469748677084025e-05,
+      "loss": 4.0214,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6461362729573505e-05,
+      "loss": 4.0283,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6452976782062985e-05,
+      "loss": 4.0217,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6444590834552465e-05,
+      "loss": 4.0335,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643620488704194e-05,
+      "loss": 4.0118,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6427835318335154e-05,
+      "loss": 4.0345,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641944937082464e-05,
+      "loss": 4.0227,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6411063423314114e-05,
+      "loss": 4.0298,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6402677475803594e-05,
+      "loss": 4.0408,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6394291528293074e-05,
+      "loss": 4.025,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638592195958629e-05,
+      "loss": 4.0309,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637753601207576e-05,
+      "loss": 4.0319,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636915006456524e-05,
+      "loss": 4.0121,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636076411705472e-05,
+      "loss": 4.0179,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63523781695442e-05,
+      "loss": 4.0196,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634400860083741e-05,
+      "loss": 4.0318,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633562265332689e-05,
+      "loss": 4.0334,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632723670581638e-05,
+      "loss": 4.0278,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631885075830586e-05,
+      "loss": 4.0271,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631048118959907e-05,
+      "loss": 4.0206,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630209524208855e-05,
+      "loss": 4.0114,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629370929457803e-05,
+      "loss": 4.0251,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628532334706751e-05,
+      "loss": 4.0314,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627693739955699e-05,
+      "loss": 4.0349,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626855145204647e-05,
+      "loss": 4.0173,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626016550453595e-05,
+      "loss": 4.0192,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.056171417236328,
+      "eval_runtime": 545.3106,
+      "eval_samples_per_second": 699.768,
+      "eval_steps_per_second": 21.868,
+      "step": 839509
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625177955702543e-05,
+      "loss": 4.0497,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624339360951491e-05,
+      "loss": 4.0253,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623500766200439e-05,
+      "loss": 4.0224,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622662171449387e-05,
+      "loss": 4.0178,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621823576698335e-05,
+      "loss": 4.0201,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620984981947283e-05,
+      "loss": 4.0118,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620146387196231e-05,
+      "loss": 4.0124,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619307792445179e-05,
+      "loss": 4.0111,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618469197694127e-05,
+      "loss": 4.0223,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.617630602943075e-05,
+      "loss": 4.0232,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6167936460723956e-05,
+      "loss": 4.019,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6159550513213436e-05,
+      "loss": 4.0243,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6151164565702916e-05,
+      "loss": 4.0273,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6142778618192396e-05,
+      "loss": 4.0178,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6134392670681876e-05,
+      "loss": 4.0287,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6126006723171356e-05,
+      "loss": 4.0078,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6117620775660836e-05,
+      "loss": 4.0062,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6109234828150316e-05,
+      "loss": 4.0291,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6100848880639796e-05,
+      "loss": 4.0211,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6092462933129276e-05,
+      "loss": 4.0263,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6084076985618756e-05,
+      "loss": 4.0171,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607569103810824e-05,
+      "loss": 4.0329,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606732146940145e-05,
+      "loss": 4.0201,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605893552189093e-05,
+      "loss": 4.022,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605054957438041e-05,
+      "loss": 4.0214,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604216362686989e-05,
+      "loss": 4.0202,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6033777679359364e-05,
+      "loss": 4.0205,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6025391731848844e-05,
+      "loss": 4.0113,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6017005784338324e-05,
+      "loss": 4.0252,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.600863621563154e-05,
+      "loss": 4.0174,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600025026812101e-05,
+      "loss": 4.0064,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.599186432061049e-05,
+      "loss": 4.0196,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598347837309998e-05,
+      "loss": 4.0181,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597509242558946e-05,
+      "loss": 4.027,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596670647807894e-05,
+      "loss": 4.014,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595832053056842e-05,
+      "loss": 4.0186,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594995096186163e-05,
+      "loss": 4.0202,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594156501435111e-05,
+      "loss": 4.0209,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593317906684059e-05,
+      "loss": 4.0298,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592479311933007e-05,
+      "loss": 4.0081,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591640717181955e-05,
+      "loss": 4.0117,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590802122430903e-05,
+      "loss": 4.0091,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589963527679851e-05,
+      "loss": 4.012,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589124932928799e-05,
+      "loss": 4.0212,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58828797605812e-05,
+      "loss": 4.0098,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5874510191874414e-05,
+      "loss": 4.0265,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5866124244363894e-05,
+      "loss": 4.0136,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5857738296853374e-05,
+      "loss": 4.0287,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5849352349342854e-05,
+      "loss": 4.0183,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5840966401832334e-05,
+      "loss": 3.9984,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583259683312554e-05,
+      "loss": 4.0123,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582421088561502e-05,
+      "loss": 4.0091,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58158249381045e-05,
+      "loss": 4.0174,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580743899059398e-05,
+      "loss": 4.0117,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579905304308346e-05,
+      "loss": 4.0123,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579066709557294e-05,
+      "loss": 3.9964,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578228114806242e-05,
+      "loss": 4.0145,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57738952005519e-05,
+      "loss": 4.0031,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576552563184512e-05,
+      "loss": 4.018,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57571396843346e-05,
+      "loss": 4.0022,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574875373682408e-05,
+      "loss": 4.0124,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574036778931355e-05,
+      "loss": 4.006,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573198184180303e-05,
+      "loss": 3.9992,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572361227309625e-05,
+      "loss": 4.0129,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571522632558573e-05,
+      "loss": 4.0099,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57068403780752e-05,
+      "loss": 4.018,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569845443056468e-05,
+      "loss": 4.0014,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569006848305416e-05,
+      "loss": 4.0111,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5681698914347376e-05,
+      "loss": 4.0026,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5673312966836856e-05,
+      "loss": 4.0091,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5664927019326336e-05,
+      "loss": 4.0082,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5656541071815816e-05,
+      "loss": 4.0172,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5648171503109025e-05,
+      "loss": 4.0064,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5639785555598505e-05,
+      "loss": 4.0312,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5631399608087985e-05,
+      "loss": 4.0218,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5623013660577465e-05,
+      "loss": 4.0103,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5614644091870674e-05,
+      "loss": 4.0138,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5606258144360154e-05,
+      "loss": 4.0052,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5597872196849634e-05,
+      "loss": 4.0065,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5589486249339114e-05,
+      "loss": 4.0148,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5581100301828594e-05,
+      "loss": 4.0199,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557273073312181e-05,
+      "loss": 4.0205,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556436116441502e-05,
+      "loss": 3.9984,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55559752169045e-05,
+      "loss": 4.0076,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554758926939398e-05,
+      "loss": 3.9971,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553920332188346e-05,
+      "loss": 4.0133,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553081737437294e-05,
+      "loss": 4.0041,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552243142686242e-05,
+      "loss": 4.0119,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55140454793519e-05,
+      "loss": 4.0192,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550565953184138e-05,
+      "loss": 4.0165,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549727358433086e-05,
+      "loss": 4.0,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548890401562407e-05,
+      "loss": 4.0052,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5480534446917277e-05,
+      "loss": 4.0146,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5472148499406757e-05,
+      "loss": 4.0004,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546376255189624e-05,
+      "loss": 3.9981,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545537660438572e-05,
+      "loss": 4.0123,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54469906568752e-05,
+      "loss": 4.0173,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543860470936468e-05,
+      "loss": 4.0228,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543021876185416e-05,
+      "loss": 4.0124,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542183281434364e-05,
+      "loss": 4.0122,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.541346324563685e-05,
+      "loss": 4.0009,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.540507729812633e-05,
+      "loss": 4.0131,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.539669135061581e-05,
+      "loss": 4.0107,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538832178190902e-05,
+      "loss": 4.0199,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537995221320223e-05,
+      "loss": 4.0098,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537156626569171e-05,
+      "loss": 3.9965,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53631803181812e-05,
+      "loss": 3.999,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5354810749474406e-05,
+      "loss": 3.9932,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5346424801963886e-05,
+      "loss": 4.007,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5338038854453366e-05,
+      "loss": 4.0127,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5329652906942846e-05,
+      "loss": 4.0107,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5321266959432326e-05,
+      "loss": 3.9948,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5312881011921806e-05,
+      "loss": 4.0013,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5304495064411286e-05,
+      "loss": 4.0142,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5296109116900766e-05,
+      "loss": 4.014,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5287723169390246e-05,
+      "loss": 4.0149,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5279337221879726e-05,
+      "loss": 4.0065,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5270951274369206e-05,
+      "loss": 3.998,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5262581705662415e-05,
+      "loss": 4.0109,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5254195758151895e-05,
+      "loss": 4.0161,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524582618944511e-05,
+      "loss": 4.0053,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523744024193459e-05,
+      "loss": 3.9971,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522905429442407e-05,
+      "loss": 4.0027,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522066834691355e-05,
+      "loss": 4.0033,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521228239940303e-05,
+      "loss": 4.0101,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.520389645189251e-05,
+      "loss": 4.0043,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5195510504381983e-05,
+      "loss": 4.0166,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51871409356752e-05,
+      "loss": 3.9959,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517877136696841e-05,
+      "loss": 4.018,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517038541945789e-05,
+      "loss": 4.0013,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516199947194737e-05,
+      "loss": 4.014,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515361352443685e-05,
+      "loss": 4.0204,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5145227576926335e-05,
+      "loss": 4.0121,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513684162941581e-05,
+      "loss": 4.0107,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512845568190529e-05,
+      "loss": 4.0154,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512006973439477e-05,
+      "loss": 3.9931,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5111700165687984e-05,
+      "loss": 4.0007,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510331421817746e-05,
+      "loss": 4.0019,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509494464947067e-05,
+      "loss": 4.0163,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508655870196015e-05,
+      "loss": 4.0152,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5078172754449626e-05,
+      "loss": 4.0145,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5069786806939106e-05,
+      "loss": 4.007,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506141723823232e-05,
+      "loss": 4.0005,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50530312907218e-05,
+      "loss": 3.9944,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504464534321128e-05,
+      "loss": 4.0082,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503625939570076e-05,
+      "loss": 4.0111,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502787344819024e-05,
+      "loss": 4.0202,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501948750067972e-05,
+      "loss": 3.9986,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50111015531692e-05,
+      "loss": 4.0046,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0454487800598145,
+      "eval_runtime": 536.0659,
+      "eval_samples_per_second": 711.836,
+      "eval_steps_per_second": 22.245,
+      "step": 915828
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500271560565868e-05,
+      "loss": 3.9986,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499432965814816e-05,
+      "loss": 4.0223,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498594371063764e-05,
+      "loss": 3.9963,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497755776312712e-05,
+      "loss": 3.9871,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49691718156166e-05,
+      "loss": 4.0089,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496078586810608e-05,
+      "loss": 4.0062,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495239992059556e-05,
+      "loss": 4.0066,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494401397308504e-05,
+      "loss": 3.9971,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493562802557452e-05,
+      "loss": 3.9998,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4927242078064e-05,
+      "loss": 3.9843,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491885613055348e-05,
+      "loss": 4.0186,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491047018304296e-05,
+      "loss": 3.9971,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490208423553244e-05,
+      "loss": 4.0116,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489369828802192e-05,
+      "loss": 4.0107,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48853123405114e-05,
+      "loss": 4.0021,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487692639300088e-05,
+      "loss": 3.9959,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486854044549036e-05,
+      "loss": 4.0076,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486017087678357e-05,
+      "loss": 4.0055,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485178492927305e-05,
+      "loss": 3.9966,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484339898176253e-05,
+      "loss": 4.0172,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483501303425201e-05,
+      "loss": 4.0224,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482662708674149e-05,
+      "loss": 4.0043,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481824113923097e-05,
+      "loss": 3.9955,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480985519172045e-05,
+      "loss": 4.0045,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4801469244209937e-05,
+      "loss": 4.0006,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4793099675503146e-05,
+      "loss": 3.9968,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4784713727992626e-05,
+      "loss": 4.0003,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4776344159285835e-05,
+      "loss": 3.9987,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4767958211775315e-05,
+      "loss": 4.003,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4759572264264795e-05,
+      "loss": 4.0073,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4751186316754275e-05,
+      "loss": 4.0084,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4742800369243755e-05,
+      "loss": 4.002,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473441442173323e-05,
+      "loss": 4.0042,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472602847422271e-05,
+      "loss": 4.0125,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471764252671219e-05,
+      "loss": 3.9925,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4709272958005403e-05,
+      "loss": 4.0003,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4700887010494883e-05,
+      "loss": 3.9981,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469250106298436e-05,
+      "loss": 4.0142,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468411511547384e-05,
+      "loss": 4.0134,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467572916796332e-05,
+      "loss": 4.012,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466735959925653e-05,
+      "loss": 4.0064,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465897365174601e-05,
+      "loss": 4.007,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465058770423549e-05,
+      "loss": 4.0002,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464220175672497e-05,
+      "loss": 3.9939,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463381580921445e-05,
+      "loss": 3.9885,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462544624050766e-05,
+      "loss": 3.994,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461706029299714e-05,
+      "loss": 3.9961,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460867434548663e-05,
+      "loss": 3.9979,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460028839797611e-05,
+      "loss": 3.9946,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459190245046559e-05,
+      "loss": 4.0027,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45835328817588e-05,
+      "loss": 3.9895,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457514693424828e-05,
+      "loss": 3.9994,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456676098673776e-05,
+      "loss": 4.0004,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455837503922724e-05,
+      "loss": 3.9807,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550005470520446e-05,
+      "loss": 3.9886,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4541619523009926e-05,
+      "loss": 3.9963,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533233575499406e-05,
+      "loss": 4.0103,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4524847627988886e-05,
+      "loss": 3.9834,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516461680478366e-05,
+      "loss": 3.9983,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450809211177158e-05,
+      "loss": 4.0015,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449970616426106e-05,
+      "loss": 3.9913,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449132021675054e-05,
+      "loss": 3.9843,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448295064804375e-05,
+      "loss": 3.9849,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447456470053323e-05,
+      "loss": 3.9853,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446617875302271e-05,
+      "loss": 4.0053,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445779280551219e-05,
+      "loss": 4.0054,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444940685800167e-05,
+      "loss": 3.9904,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444102091049115e-05,
+      "loss": 3.9983,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443265134178436e-05,
+      "loss": 4.0058,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442426539427384e-05,
+      "loss": 3.9841,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441587944676332e-05,
+      "loss": 4.0105,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44074934992528e-05,
+      "loss": 3.9967,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439910755174228e-05,
+      "loss": 3.9924,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439072160423176e-05,
+      "loss": 3.9906,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4382335656721246e-05,
+      "loss": 4.0079,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437394970921072e-05,
+      "loss": 3.9902,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4365580140503935e-05,
+      "loss": 4.006,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4357194192993415e-05,
+      "loss": 3.9996,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4348824624286624e-05,
+      "loss": 3.996,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340438676776104e-05,
+      "loss": 3.9955,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433206910806931e-05,
+      "loss": 3.998,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432368316055879e-05,
+      "loss": 3.99,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431529721304827e-05,
+      "loss": 3.9907,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430691126553775e-05,
+      "loss": 3.9945,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429852531802723e-05,
+      "loss": 3.9782,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429013937051671e-05,
+      "loss": 4.0,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428175342300619e-05,
+      "loss": 3.984,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427336747549567e-05,
+      "loss": 3.9949,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426498152798515e-05,
+      "loss": 3.9825,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425661195927837e-05,
+      "loss": 4.0052,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424822601176784e-05,
+      "loss": 3.9983,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423984006425732e-05,
+      "loss": 4.0023,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.42314541167468e-05,
+      "loss": 4.0019,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422306816923628e-05,
+      "loss": 3.9818,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421468222172576e-05,
+      "loss": 3.9985,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420629627421524e-05,
+      "loss": 3.9947,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419791032670472e-05,
+      "loss": 3.9896,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418954075799794e-05,
+      "loss": 3.9852,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4181171189291147e-05,
+      "loss": 3.9885,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4172785241780627e-05,
+      "loss": 4.0045,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4164399294270107e-05,
+      "loss": 3.9889,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156013346759586e-05,
+      "loss": 4.0029,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147643778052796e-05,
+      "loss": 3.9987,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139257830542276e-05,
+      "loss": 3.9932,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4130871883031755e-05,
+      "loss": 3.9776,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122485935521235e-05,
+      "loss": 3.9911,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114099988010715e-05,
+      "loss": 3.9975,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4105714040500195e-05,
+      "loss": 4.0007,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4097328092989675e-05,
+      "loss": 3.9887,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4088942145479155e-05,
+      "loss": 3.9828,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408057257677237e-05,
+      "loss": 3.992,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407218662926185e-05,
+      "loss": 3.9971,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406380068175133e-05,
+      "loss": 3.9859,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405541473424081e-05,
+      "loss": 3.9907,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404702878673029e-05,
+      "loss": 3.9869,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403864283921977e-05,
+      "loss": 3.9839,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403027327051298e-05,
+      "loss": 3.9878,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402188732300246e-05,
+      "loss": 3.9925,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401350137549194e-05,
+      "loss": 3.9939,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400511542798142e-05,
+      "loss": 3.9876,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39967294804709e-05,
+      "loss": 4.0012,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398834353296037e-05,
+      "loss": 4.009,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397995758544986e-05,
+      "loss": 3.988,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3971588016743076e-05,
+      "loss": 3.9987,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3963202069232556e-05,
+      "loss": 3.989,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395481612172203e-05,
+      "loss": 3.9929,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394643017421151e-05,
+      "loss": 3.9971,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393804422670099e-05,
+      "loss": 3.9873,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392965827919047e-05,
+      "loss": 3.9935,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392127233167995e-05,
+      "loss": 3.9846,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391288638416943e-05,
+      "loss": 4.0055,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.390451681546264e-05,
+      "loss": 3.9866,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.389613086795212e-05,
+      "loss": 4.0045,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38877449204416e-05,
+      "loss": 3.9927,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387935897293108e-05,
+      "loss": 3.9985,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387097302542056e-05,
+      "loss": 3.9918,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386260345671377e-05,
+      "loss": 3.9924,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385421750920325e-05,
+      "loss": 4.0031,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384584794049646e-05,
+      "loss": 3.9887,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383746199298594e-05,
+      "loss": 3.9961,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382907604547542e-05,
+      "loss": 3.9984,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38206900979649e-05,
+      "loss": 3.9882,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381230415045438e-05,
+      "loss": 3.9867,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380391820294386e-05,
+      "loss": 3.9957,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379553225543334e-05,
+      "loss": 3.9929,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378714630792282e-05,
+      "loss": 3.9942,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37787603604123e-05,
+      "loss": 3.9933,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377039079170551e-05,
+      "loss": 3.9918,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376200484419499e-05,
+      "loss": 3.9833,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.03564453125,
+      "eval_runtime": 302.9696,
+      "eval_samples_per_second": 1259.503,
+      "eval_steps_per_second": 39.36,
+      "step": 992148
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375361889668448e-05,
+      "loss": 3.9887,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374523294917396e-05,
+      "loss": 4.008,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373686338046717e-05,
+      "loss": 3.9855,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372847743295665e-05,
+      "loss": 3.9718,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372009148544613e-05,
+      "loss": 3.9943,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371170553793561e-05,
+      "loss": 3.9877,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370331959042509e-05,
+      "loss": 3.9931,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369493364291456e-05,
+      "loss": 3.9809,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368654769540404e-05,
+      "loss": 3.9838,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3678178126697256e-05,
+      "loss": 3.9698,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3669792179186736e-05,
+      "loss": 4.0009,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3661406231676216e-05,
+      "loss": 3.9827,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3653020284165696e-05,
+      "loss": 3.9914,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3644634336655176e-05,
+      "loss": 3.9985,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363626476794839e-05,
+      "loss": 3.9855,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3627878820437865e-05,
+      "loss": 3.9742,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3619492872927345e-05,
+      "loss": 3.9978,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3611106925416824e-05,
+      "loss": 3.988,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3602737356710034e-05,
+      "loss": 3.981,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3594351409199514e-05,
+      "loss": 4.0014,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358598184049273e-05,
+      "loss": 4.0064,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357759589298221e-05,
+      "loss": 3.9901,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356920994547168e-05,
+      "loss": 3.9796,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35608403767649e-05,
+      "loss": 3.9873,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3552454429254385e-05,
+      "loss": 3.9833,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354406848174386e-05,
+      "loss": 3.9816,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353568253423334e-05,
+      "loss": 3.9868,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352729658672282e-05,
+      "loss": 3.9821,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35189106392123e-05,
+      "loss": 3.9867,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351052469170178e-05,
+      "loss": 3.9915,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350213874419126e-05,
+      "loss": 3.9925,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349375279668074e-05,
+      "loss": 3.9874,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.348538322797395e-05,
+      "loss": 3.9874,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347699728046343e-05,
+      "loss": 3.9966,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346861133295291e-05,
+      "loss": 3.9789,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346022538544239e-05,
+      "loss": 3.9864,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345183943793187e-05,
+      "loss": 3.9808,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3443453490421354e-05,
+      "loss": 3.9945,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3435067542910834e-05,
+      "loss": 3.9961,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342669797420404e-05,
+      "loss": 4.0013,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341831202669352e-05,
+      "loss": 3.9919,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3409926079183e-05,
+      "loss": 3.9928,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340154013167248e-05,
+      "loss": 3.9813,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339317056296569e-05,
+      "loss": 3.9795,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338478461545517e-05,
+      "loss": 3.9758,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337639866794465e-05,
+      "loss": 3.9737,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336801272043413e-05,
+      "loss": 3.9764,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335962677292361e-05,
+      "loss": 3.9839,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335124082541309e-05,
+      "loss": 3.984,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334285487790257e-05,
+      "loss": 3.9841,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333448530919579e-05,
+      "loss": 3.9724,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332609936168527e-05,
+      "loss": 3.9836,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331771341417475e-05,
+      "loss": 3.9835,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330932746666422e-05,
+      "loss": 3.9679,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33009415191537e-05,
+      "loss": 3.9723,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.329255557164318e-05,
+      "loss": 3.9807,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.328416962413266e-05,
+      "loss": 3.9936,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.327578367662214e-05,
+      "loss": 3.9645,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326741410791535e-05,
+      "loss": 3.9842,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325902816040483e-05,
+      "loss": 3.9886,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325064221289431e-05,
+      "loss": 3.9745,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324225626538379e-05,
+      "loss": 3.9681,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323387031787327e-05,
+      "loss": 3.9677,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3225500749166485e-05,
+      "loss": 3.9697,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3217131180459694e-05,
+      "loss": 3.9884,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320876161175291e-05,
+      "loss": 3.9926,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320037566424239e-05,
+      "loss": 3.9762,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319198971673187e-05,
+      "loss": 3.9834,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318360376922134e-05,
+      "loss": 3.9858,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317521782171082e-05,
+      "loss": 3.9754,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31668318742003e-05,
+      "loss": 3.9912,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315844592668978e-05,
+      "loss": 3.9784,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315005997917926e-05,
+      "loss": 3.9798,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314167403166874e-05,
+      "loss": 3.9741,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313330446296196e-05,
+      "loss": 3.9904,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312491851545144e-05,
+      "loss": 3.9794,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311653256794092e-05,
+      "loss": 3.989,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31081466204304e-05,
+      "loss": 3.9837,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309976067291988e-05,
+      "loss": 3.9823,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309137472540936e-05,
+      "loss": 3.9794,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308298877789884e-05,
+      "loss": 3.9836,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.307461920919205e-05,
+      "loss": 3.974,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306623326168153e-05,
+      "loss": 3.9761,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305784731417101e-05,
+      "loss": 3.9767,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304946136666049e-05,
+      "loss": 3.9665,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304107541914997e-05,
+      "loss": 3.9819,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303268947163945e-05,
+      "loss": 3.9689,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302430352412893e-05,
+      "loss": 3.9805,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301591757661841e-05,
+      "loss": 3.9659,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.300756438671535e-05,
+      "loss": 3.9877,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299917843920483e-05,
+      "loss": 3.9846,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299079249169431e-05,
+      "loss": 3.9872,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298240654418379e-05,
+      "loss": 3.9839,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2974036975477e-05,
+      "loss": 3.9705,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296565102796648e-05,
+      "loss": 3.9839,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295726508045596e-05,
+      "loss": 3.9775,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294887913294544e-05,
+      "loss": 3.9762,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294049318543492e-05,
+      "loss": 3.9682,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29321072379244e-05,
+      "loss": 3.9725,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292373766921762e-05,
+      "loss": 3.9864,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29153517217071e-05,
+      "loss": 3.9784,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290696577419658e-05,
+      "loss": 3.9881,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289857982668606e-05,
+      "loss": 3.9791,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2890210257979266e-05,
+      "loss": 3.978,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2881824310468746e-05,
+      "loss": 3.9687,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2873438362958226e-05,
+      "loss": 3.9709,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2865052415447706e-05,
+      "loss": 3.984,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285666646793718e-05,
+      "loss": 3.9833,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284828052042666e-05,
+      "loss": 3.9735,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283989457291614e-05,
+      "loss": 3.9667,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283150862540562e-05,
+      "loss": 3.9797,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28231226778951e-05,
+      "loss": 3.9812,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281476948799205e-05,
+      "loss": 3.9686,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280638354048153e-05,
+      "loss": 3.981,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2797997592971004e-05,
+      "loss": 3.9693,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2789611645460484e-05,
+      "loss": 3.9728,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2781225697949964e-05,
+      "loss": 3.9682,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2772839750439444e-05,
+      "loss": 3.9818,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2764453802928923e-05,
+      "loss": 3.9797,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2756067855418403e-05,
+      "loss": 3.9732,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274768190790788e-05,
+      "loss": 3.9843,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273929596039736e-05,
+      "loss": 3.9935,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273091001288684e-05,
+      "loss": 3.9783,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272252406537632e-05,
+      "loss": 3.9783,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271415449666953e-05,
+      "loss": 3.9765,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270576854915902e-05,
+      "loss": 3.9789,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26973826016485e-05,
+      "loss": 3.9779,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268901303294171e-05,
+      "loss": 3.9709,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268062708543119e-05,
+      "loss": 3.9796,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267224113792067e-05,
+      "loss": 3.9723,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266385519041015e-05,
+      "loss": 3.9929,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265546924289963e-05,
+      "loss": 3.972,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264708329538911e-05,
+      "loss": 3.9902,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263869734787859e-05,
+      "loss": 3.9764,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263031140036807e-05,
+      "loss": 3.9815,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262192545285754e-05,
+      "loss": 3.9769,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261353950534702e-05,
+      "loss": 3.9828,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26051535578365e-05,
+      "loss": 3.9874,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259678398912972e-05,
+      "loss": 3.975,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25883980416192e-05,
+      "loss": 3.9814,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258001209410868e-05,
+      "loss": 3.9817,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257162614659816e-05,
+      "loss": 3.9724,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2563256577891366e-05,
+      "loss": 3.975,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2554870630380846e-05,
+      "loss": 3.9817,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2546484682870326e-05,
+      "loss": 3.9793,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2538098735359806e-05,
+      "loss": 3.98,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2529712787849286e-05,
+      "loss": 3.9795,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2521343219142495e-05,
+      "loss": 3.9753,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2512957271631975e-05,
+      "loss": 3.9707,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.027347087860107,
+      "eval_runtime": 293.1436,
+      "eval_samples_per_second": 1301.72,
+      "eval_steps_per_second": 40.68,
+      "step": 1068468
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2504571324121455e-05,
+      "loss": 3.961,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249618537661094e-05,
+      "loss": 3.9639,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248779942910042e-05,
+      "loss": 3.9738,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24794134815899e-05,
+      "loss": 3.986,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247102753407938e-05,
+      "loss": 3.9781,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246264158656886e-05,
+      "loss": 3.9765,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245425563905834e-05,
+      "loss": 3.9733,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.244586969154782e-05,
+      "loss": 3.9757,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24374837440373e-05,
+      "loss": 3.9629,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242909779652678e-05,
+      "loss": 3.9735,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2420711849016254e-05,
+      "loss": 3.9744,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2412325901505734e-05,
+      "loss": 3.9891,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240395633279895e-05,
+      "loss": 3.9875,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239557038528842e-05,
+      "loss": 3.9546,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238718443777791e-05,
+      "loss": 3.9777,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237879849026739e-05,
+      "loss": 3.9621,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237041254275687e-05,
+      "loss": 3.9775,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236202659524635e-05,
+      "loss": 3.9716,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235364064773583e-05,
+      "loss": 3.977,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234525470022531e-05,
+      "loss": 3.9687,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.233686875271479e-05,
+      "loss": 3.9914,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232848280520427e-05,
+      "loss": 3.9717,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232009685769375e-05,
+      "loss": 3.978,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231171091018323e-05,
+      "loss": 3.9711,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230332496267271e-05,
+      "loss": 3.9834,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229497177276965e-05,
+      "loss": 3.9721,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228658582525913e-05,
+      "loss": 3.965,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227819987774861e-05,
+      "loss": 3.9689,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226981393023809e-05,
+      "loss": 3.9732,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2261427982727574e-05,
+      "loss": 3.968,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225305841402078e-05,
+      "loss": 3.9699,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224467246651026e-05,
+      "loss": 3.9721,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223628651899974e-05,
+      "loss": 3.9806,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222790057148922e-05,
+      "loss": 3.9777,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22195146239787e-05,
+      "loss": 3.9799,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221116143407564e-05,
+      "loss": 3.9759,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220277548656512e-05,
+      "loss": 3.982,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21943895390546e-05,
+      "loss": 3.9607,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218600359154408e-05,
+      "loss": 3.967,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217761764403356e-05,
+      "loss": 3.9699,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216923169652304e-05,
+      "loss": 3.9603,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216084574901253e-05,
+      "loss": 3.9746,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215245980150201e-05,
+      "loss": 3.9722,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214407385399149e-05,
+      "loss": 3.9667,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213568790648097e-05,
+      "loss": 3.9784,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212730195897044e-05,
+      "loss": 3.9756,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211891601145992e-05,
+      "loss": 3.9758,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211054644275314e-05,
+      "loss": 3.9741,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210216049524262e-05,
+      "loss": 3.9555,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2093790926535826e-05,
+      "loss": 3.972,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2085421357829035e-05,
+      "loss": 3.973,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2077035410318515e-05,
+      "loss": 3.9783,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2068649462807995e-05,
+      "loss": 3.9683,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206026351529748e-05,
+      "loss": 3.9636,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205187756778696e-05,
+      "loss": 3.963,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204349162027644e-05,
+      "loss": 3.9615,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2035105672765915e-05,
+      "loss": 3.9714,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2026719725255395e-05,
+      "loss": 3.9515,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2018333777744875e-05,
+      "loss": 3.98,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2009947830234355e-05,
+      "loss": 3.9751,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2001561882723834e-05,
+      "loss": 3.9511,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1993175935213314e-05,
+      "loss": 3.9826,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1984806366506524e-05,
+      "loss": 3.9558,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1976420418996004e-05,
+      "loss": 3.9624,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196805085028922e-05,
+      "loss": 3.9626,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19596649027787e-05,
+      "loss": 3.9705,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195127895526818e-05,
+      "loss": 3.9592,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.194289300775766e-05,
+      "loss": 3.9768,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193452343905087e-05,
+      "loss": 3.9494,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192613749154035e-05,
+      "loss": 3.9611,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191775154402983e-05,
+      "loss": 3.9748,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190936559651931e-05,
+      "loss": 3.9664,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190097964900879e-05,
+      "loss": 3.976,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189259370149827e-05,
+      "loss": 3.9719,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188420775398775e-05,
+      "loss": 3.9825,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187582180647723e-05,
+      "loss": 3.9608,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186743585896671e-05,
+      "loss": 3.9814,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185904991145619e-05,
+      "loss": 3.9623,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185066396394567e-05,
+      "loss": 3.9622,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184227801643515e-05,
+      "loss": 3.9659,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183389206892463e-05,
+      "loss": 3.9761,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1825522500217844e-05,
+      "loss": 3.9686,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181715293151105e-05,
+      "loss": 3.9811,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180876698400053e-05,
+      "loss": 3.958,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180038103649001e-05,
+      "loss": 3.9528,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179199508897949e-05,
+      "loss": 3.9616,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178360914146897e-05,
+      "loss": 3.9694,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177523957276218e-05,
+      "loss": 3.9654,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176687000405539e-05,
+      "loss": 3.9654,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.175848405654487e-05,
+      "loss": 3.9667,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175009810903436e-05,
+      "loss": 3.9641,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174171216152384e-05,
+      "loss": 3.9644,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173332621401332e-05,
+      "loss": 3.9615,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.17249402665028e-05,
+      "loss": 3.9768,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171655431899228e-05,
+      "loss": 3.9793,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170816837148175e-05,
+      "loss": 3.9719,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169978242397123e-05,
+      "loss": 3.9588,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169139647646071e-05,
+      "loss": 3.9635,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168301052895019e-05,
+      "loss": 3.9634,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167462458143967e-05,
+      "loss": 3.9736,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166625501273288e-05,
+      "loss": 3.9758,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165786906522236e-05,
+      "loss": 3.96,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164948311771184e-05,
+      "loss": 3.9687,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164109717020132e-05,
+      "loss": 3.9667,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1632711222690806e-05,
+      "loss": 3.9678,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1624341653984015e-05,
+      "loss": 3.9615,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1615972085277224e-05,
+      "loss": 3.9623,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1607586137766704e-05,
+      "loss": 3.9508,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1599200190256184e-05,
+      "loss": 3.966,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1590814242745664e-05,
+      "loss": 3.9573,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1582428295235144e-05,
+      "loss": 3.9629,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1574042347724624e-05,
+      "loss": 3.9613,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1565656400214104e-05,
+      "loss": 3.9637,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1557270452703584e-05,
+      "loss": 3.9607,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154890088399679e-05,
+      "loss": 3.9693,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154051493648627e-05,
+      "loss": 3.9709,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153212898897576e-05,
+      "loss": 3.9739,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152374304146524e-05,
+      "loss": 3.9733,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151535709395472e-05,
+      "loss": 3.9613,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150698752524793e-05,
+      "loss": 3.9664,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149861795654114e-05,
+      "loss": 3.9676,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149023200903062e-05,
+      "loss": 3.9507,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14818460615201e-05,
+      "loss": 3.9678,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.147346011400958e-05,
+      "loss": 3.9588,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146507416649906e-05,
+      "loss": 3.9719,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145668821898854e-05,
+      "loss": 3.9612,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144830227147802e-05,
+      "loss": 3.97,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14399163239675e-05,
+      "loss": 3.9591,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143153037645698e-05,
+      "loss": 3.9589,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142314442894646e-05,
+      "loss": 3.9708,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141477486023967e-05,
+      "loss": 3.9694,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140638891272915e-05,
+      "loss": 3.9753,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139800296521863e-05,
+      "loss": 3.959,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138961701770811e-05,
+      "loss": 3.9746,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1381231070197586e-05,
+      "loss": 3.9681,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13728615014908e-05,
+      "loss": 3.9669,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136449193278401e-05,
+      "loss": 3.973,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135610598527349e-05,
+      "loss": 3.9614,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13477364165667e-05,
+      "loss": 3.9666,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133935046905618e-05,
+      "loss": 3.9736,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133096452154567e-05,
+      "loss": 3.9564,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132257857403515e-05,
+      "loss": 3.9572,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131419262652463e-05,
+      "loss": 3.9673,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130580667901411e-05,
+      "loss": 3.9655,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129742073150359e-05,
+      "loss": 3.963,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128903478399306e-05,
+      "loss": 3.9689,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128064883648254e-05,
+      "loss": 3.9604,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127226288897202e-05,
+      "loss": 3.9814,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.12638769414615e-05,
+      "loss": 3.9639,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.019488334655762,
+      "eval_runtime": 484.9846,
+      "eval_samples_per_second": 786.81,
+      "eval_steps_per_second": 24.588,
+      "step": 1144788
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.125550737275471e-05,
+      "loss": 3.9595,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124712142524419e-05,
+      "loss": 3.9579,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123873547773367e-05,
+      "loss": 3.9535,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123034953022315e-05,
+      "loss": 3.977,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1221963582712636e-05,
+      "loss": 3.9664,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1213577635202116e-05,
+      "loss": 3.9683,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1205191687691595e-05,
+      "loss": 3.9591,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1196805740181075e-05,
+      "loss": 3.962,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1188419792670555e-05,
+      "loss": 3.9535,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1180050223963764e-05,
+      "loss": 3.9606,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1171664276453244e-05,
+      "loss": 3.962,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1163294707746454e-05,
+      "loss": 3.9786,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115492513903966e-05,
+      "loss": 3.9731,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114653919152914e-05,
+      "loss": 3.9437,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113815324401862e-05,
+      "loss": 3.9631,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11297672965081e-05,
+      "loss": 3.9528,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112138134899759e-05,
+      "loss": 3.9619,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.111299540148707e-05,
+      "loss": 3.9593,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.110460945397655e-05,
+      "loss": 3.963,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.109622350646603e-05,
+      "loss": 3.9589,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108783755895551e-05,
+      "loss": 3.9761,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107945161144499e-05,
+      "loss": 3.9638,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107106566393447e-05,
+      "loss": 3.9598,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.106267971642394e-05,
+      "loss": 3.9592,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105431014771716e-05,
+      "loss": 3.9736,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.104592420020664e-05,
+      "loss": 3.9594,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103753825269612e-05,
+      "loss": 3.9536,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102915230518559e-05,
+      "loss": 3.9607,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102078273647881e-05,
+      "loss": 3.9573,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.101239678896829e-05,
+      "loss": 3.9561,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.100401084145777e-05,
+      "loss": 3.9601,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099562489394725e-05,
+      "loss": 3.9552,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098723894643673e-05,
+      "loss": 3.9665,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097885299892621e-05,
+      "loss": 3.9674,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097049980902315e-05,
+      "loss": 3.9651,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096213024031636e-05,
+      "loss": 3.9649,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095374429280584e-05,
+      "loss": 3.9676,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.094535834529532e-05,
+      "loss": 3.9482,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09369723977848e-05,
+      "loss": 3.9559,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092858645027428e-05,
+      "loss": 3.9569,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092020050276376e-05,
+      "loss": 3.9504,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091181455525324e-05,
+      "loss": 3.9601,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090342860774272e-05,
+      "loss": 3.9609,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08950426602322e-05,
+      "loss": 3.9517,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088665671272168e-05,
+      "loss": 3.9643,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087827076521116e-05,
+      "loss": 3.9646,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086988481770064e-05,
+      "loss": 3.9636,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086149887019012e-05,
+      "loss": 3.9634,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085312930148333e-05,
+      "loss": 3.9448,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084474335397281e-05,
+      "loss": 3.9604,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083637378526602e-05,
+      "loss": 3.9654,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0828004216559234e-05,
+      "loss": 3.9623,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0819618269048714e-05,
+      "loss": 3.9558,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0811232321538194e-05,
+      "loss": 3.953,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0802846374027674e-05,
+      "loss": 3.9468,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0794460426517154e-05,
+      "loss": 3.9485,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0786074479006634e-05,
+      "loss": 3.96,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0777688531496114e-05,
+      "loss": 3.9424,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0769302583985594e-05,
+      "loss": 3.9645,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07609330152788e-05,
+      "loss": 3.9624,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075254706776828e-05,
+      "loss": 3.9472,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074416112025776e-05,
+      "loss": 3.9693,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073577517274724e-05,
+      "loss": 3.9446,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072740560404045e-05,
+      "loss": 3.9486,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071901965652993e-05,
+      "loss": 3.9506,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071063370901941e-05,
+      "loss": 3.9613,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07022477615089e-05,
+      "loss": 3.9453,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.069386181399838e-05,
+      "loss": 3.9625,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.068549224529159e-05,
+      "loss": 3.9379,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067710629778107e-05,
+      "loss": 3.9484,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066872035027055e-05,
+      "loss": 3.9659,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066033440276003e-05,
+      "loss": 3.9562,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065194845524951e-05,
+      "loss": 3.9615,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.064356250773899e-05,
+      "loss": 3.9624,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063517656022847e-05,
+      "loss": 3.9697,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062679061271795e-05,
+      "loss": 3.9521,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0618421044011157e-05,
+      "loss": 3.9688,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0610035096500637e-05,
+      "loss": 3.9499,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0601649148990116e-05,
+      "loss": 3.952,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0593263201479596e-05,
+      "loss": 3.9543,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0584877253969076e-05,
+      "loss": 3.9599,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057650768526229e-05,
+      "loss": 3.9596,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056812173775177e-05,
+      "loss": 3.9675,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055973579024125e-05,
+      "loss": 3.9503,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0551349842730725e-05,
+      "loss": 3.94,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0542963895220205e-05,
+      "loss": 3.9471,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0534577947709685e-05,
+      "loss": 3.9571,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05262083790029e-05,
+      "loss": 3.954,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0517822431492374e-05,
+      "loss": 3.9523,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509436483981858e-05,
+      "loss": 3.9592,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0501050536471338e-05,
+      "loss": 3.9494,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0492664588960818e-05,
+      "loss": 3.9525,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0484295020254027e-05,
+      "loss": 3.9481,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0475909072743507e-05,
+      "loss": 3.9636,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046752312523299e-05,
+      "loss": 3.9683,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045913717772247e-05,
+      "loss": 3.9631,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045075123021195e-05,
+      "loss": 3.9444,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.044236528270143e-05,
+      "loss": 3.9547,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.043397933519091e-05,
+      "loss": 3.9517,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.042559338768039e-05,
+      "loss": 3.9593,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.04172238189736e-05,
+      "loss": 3.9641,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040885425026681e-05,
+      "loss": 3.9466,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040046830275629e-05,
+      "loss": 3.9574,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.039208235524577e-05,
+      "loss": 3.9553,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.038369640773525e-05,
+      "loss": 3.9558,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.037531046022473e-05,
+      "loss": 3.9502,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0366940891517944e-05,
+      "loss": 3.9489,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0358554944007424e-05,
+      "loss": 3.9435,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0350168996496904e-05,
+      "loss": 3.9557,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0341783048986384e-05,
+      "loss": 3.9455,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0333397101475863e-05,
+      "loss": 3.9474,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0325011153965343e-05,
+      "loss": 3.9512,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0316625206454823e-05,
+      "loss": 3.958,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0308239258944303e-05,
+      "loss": 3.9501,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0299869690237516e-05,
+      "loss": 3.9535,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0291483742726996e-05,
+      "loss": 3.9646,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0283097795216476e-05,
+      "loss": 3.9616,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0274711847705956e-05,
+      "loss": 3.9605,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0266342278999165e-05,
+      "loss": 3.9482,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0257956331488645e-05,
+      "loss": 3.9552,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0249570383978125e-05,
+      "loss": 3.9536,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0241200815271337e-05,
+      "loss": 3.9429,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0232814867760817e-05,
+      "loss": 3.955,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0224428920250297e-05,
+      "loss": 3.9482,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0216042972739777e-05,
+      "loss": 3.9587,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0207657025229257e-05,
+      "loss": 3.9497,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0199271077718737e-05,
+      "loss": 3.9576,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0190885130208214e-05,
+      "loss": 3.9488,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0182499182697693e-05,
+      "loss": 3.949,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0174113235187173e-05,
+      "loss": 3.9556,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0165743666480383e-05,
+      "loss": 3.9582,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0157357718969862e-05,
+      "loss": 3.9624,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0148971771459346e-05,
+      "loss": 3.9516,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0140585823948826e-05,
+      "loss": 3.9582,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0132199876438306e-05,
+      "loss": 3.9595,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0123830307731515e-05,
+      "loss": 3.9577,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115444360220995e-05,
+      "loss": 3.9571,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010707479151421e-05,
+      "loss": 3.9524,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098688844003687e-05,
+      "loss": 3.9547,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0090302896493167e-05,
+      "loss": 3.9626,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0081916948982647e-05,
+      "loss": 3.9489,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073531001472127e-05,
+      "loss": 3.9458,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0065145053961607e-05,
+      "loss": 3.9573,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0056759106451087e-05,
+      "loss": 3.9536,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0048373158940567e-05,
+      "loss": 3.9513,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.004000359023378e-05,
+      "loss": 3.9601,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.003163402152699e-05,
+      "loss": 3.9494,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.002324807401647e-05,
+      "loss": 3.9657,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.001486212650595e-05,
+      "loss": 3.9539,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0138750076293945,
+      "eval_runtime": 439.8499,
+      "eval_samples_per_second": 867.548,
+      "eval_steps_per_second": 27.112,
+      "step": 1221108
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.000647617899543e-05,
+      "loss": 3.9317,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.999809023148491e-05,
+      "loss": 3.945,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989704283974392e-05,
+      "loss": 3.9413,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.998131833646387e-05,
+      "loss": 3.9637,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.997293238895335e-05,
+      "loss": 3.9579,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.996454644144283e-05,
+      "loss": 3.9561,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.995616049393231e-05,
+      "loss": 3.9498,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.994777454642179e-05,
+      "loss": 3.9495,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.993938859891127e-05,
+      "loss": 3.9397,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931002651400748e-05,
+      "loss": 3.9521,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922616703890228e-05,
+      "loss": 3.9488,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914230756379708e-05,
+      "loss": 3.9689,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9905861187672924e-05,
+      "loss": 3.9634,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9897475240162397e-05,
+      "loss": 3.9329,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889089292651877e-05,
+      "loss": 3.949,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.988070334514136e-05,
+      "loss": 3.9458,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.987231739763084e-05,
+      "loss": 3.9455,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.986393145012032e-05,
+      "loss": 3.9522,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98555455026098e-05,
+      "loss": 3.9519,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984715955509928e-05,
+      "loss": 3.9462,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983877360758876e-05,
+      "loss": 3.9663,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983038766007824e-05,
+      "loss": 3.9555,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.982200171256772e-05,
+      "loss": 3.9491,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98136157650572e-05,
+      "loss": 3.9461,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.980522981754668e-05,
+      "loss": 3.9672,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9796860248839892e-05,
+      "loss": 3.9492,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788474301329372e-05,
+      "loss": 3.9439,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9780088353818852e-05,
+      "loss": 3.9495,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9771702406308332e-05,
+      "loss": 3.9441,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.976333283760154e-05,
+      "loss": 3.9457,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9754946890091025e-05,
+      "loss": 3.9482,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9746560942580505e-05,
+      "loss": 3.9469,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9738174995069985e-05,
+      "loss": 3.9557,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9729789047559458e-05,
+      "loss": 3.9521,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9721403100048938e-05,
+      "loss": 3.9563,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9713033531342154e-05,
+      "loss": 3.9532,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9704647583831634e-05,
+      "loss": 3.9569,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.969626163632111e-05,
+      "loss": 3.9398,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968787568881059e-05,
+      "loss": 3.9475,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967948974130007e-05,
+      "loss": 3.9423,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967110379378955e-05,
+      "loss": 3.9457,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.966271784627903e-05,
+      "loss": 3.945,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.965433189876851e-05,
+      "loss": 3.9518,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9645962330061722e-05,
+      "loss": 3.9407,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963759276135493e-05,
+      "loss": 3.9533,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962920681384441e-05,
+      "loss": 3.9503,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962082086633389e-05,
+      "loss": 3.9557,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961243491882337e-05,
+      "loss": 3.9517,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9604048971312855e-05,
+      "loss": 3.931,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9595679402606064e-05,
+      "loss": 3.9447,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9587293455095544e-05,
+      "loss": 3.9579,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9578907507585024e-05,
+      "loss": 3.953,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9570521560074504e-05,
+      "loss": 3.9407,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562151991367716e-05,
+      "loss": 3.9437,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9553766043857196e-05,
+      "loss": 3.9396,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545380096346676e-05,
+      "loss": 3.9342,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9536994148836156e-05,
+      "loss": 3.9554,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528608201325636e-05,
+      "loss": 3.9314,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520222253815116e-05,
+      "loss": 3.9509,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9511836306304596e-05,
+      "loss": 3.9508,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503450358794076e-05,
+      "loss": 3.9356,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.949508079008729e-05,
+      "loss": 3.956,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486694842576768e-05,
+      "loss": 3.9345,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478308895066248e-05,
+      "loss": 3.9407,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9469922947555728e-05,
+      "loss": 3.934,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9461553378848937e-05,
+      "loss": 3.9524,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9453167431338417e-05,
+      "loss": 3.9393,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.94447814838279e-05,
+      "loss": 3.9445,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.943641191512111e-05,
+      "loss": 3.9312,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942802596761059e-05,
+      "loss": 3.9338,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941964002010007e-05,
+      "loss": 3.9533,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941125407258955e-05,
+      "loss": 3.9511,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.940286812507903e-05,
+      "loss": 3.9456,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394498556372242e-05,
+      "loss": 3.9495,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386112608861722e-05,
+      "loss": 3.96,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377726661351202e-05,
+      "loss": 3.9408,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369340713840682e-05,
+      "loss": 3.9583,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9360954766330162e-05,
+      "loss": 3.9363,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9352568818819642e-05,
+      "loss": 3.9429,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.934418287130912e-05,
+      "loss": 3.9437,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9335796923798598e-05,
+      "loss": 3.9499,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9327427355091814e-05,
+      "loss": 3.9462,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9319041407581294e-05,
+      "loss": 3.9602,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310655460070767e-05,
+      "loss": 3.9419,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302269512560247e-05,
+      "loss": 3.927,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9293899943853463e-05,
+      "loss": 3.9367,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9285513996342946e-05,
+      "loss": 3.9465,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.927712804883242e-05,
+      "loss": 3.9407,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.92687421013219e-05,
+      "loss": 3.9411,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.926035615381138e-05,
+      "loss": 3.9493,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.925197020630086e-05,
+      "loss": 3.9402,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.924358425879034e-05,
+      "loss": 3.941,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235198311279823e-05,
+      "loss": 3.9393,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9226828742573032e-05,
+      "loss": 3.9533,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9218442795062512e-05,
+      "loss": 3.9552,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9210056847551992e-05,
+      "loss": 3.9508,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9201670900041472e-05,
+      "loss": 3.9343,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9193301331334684e-05,
+      "loss": 3.9415,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9184915383824164e-05,
+      "loss": 3.9448,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9176529436313644e-05,
+      "loss": 3.9475,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9168143488803124e-05,
+      "loss": 3.9536,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9159757541292604e-05,
+      "loss": 3.9392,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9151387972585813e-05,
+      "loss": 3.9427,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143002025075293e-05,
+      "loss": 3.9464,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9134616077564773e-05,
+      "loss": 3.948,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126230130054256e-05,
+      "loss": 3.9351,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9117860561347466e-05,
+      "loss": 3.9427,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9109474613836945e-05,
+      "loss": 3.9333,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9101105045130155e-05,
+      "loss": 3.9375,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092719097619635e-05,
+      "loss": 3.9383,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9084333150109118e-05,
+      "loss": 3.9365,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9075947202598598e-05,
+      "loss": 3.934,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067561255088078e-05,
+      "loss": 3.946,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059175307577558e-05,
+      "loss": 3.9411,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9050789360067038e-05,
+      "loss": 3.9406,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9042403412556518e-05,
+      "loss": 3.9518,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9034033843849727e-05,
+      "loss": 3.9536,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.902564789633921e-05,
+      "loss": 3.9487,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901726194882869e-05,
+      "loss": 3.9399,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900887600131817e-05,
+      "loss": 3.9444,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900049005380765e-05,
+      "loss": 3.9428,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.899210410629713e-05,
+      "loss": 3.9357,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898375091639407e-05,
+      "loss": 3.9415,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897536496888355e-05,
+      "loss": 3.9386,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.896697902137303e-05,
+      "loss": 3.9456,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.895859307386251e-05,
+      "loss": 3.9385,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.895020712635199e-05,
+      "loss": 3.9478,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.894182117884147e-05,
+      "loss": 3.9439,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.893343523133095e-05,
+      "loss": 3.9404,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925049283820428e-05,
+      "loss": 3.9362,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916679715113644e-05,
+      "loss": 3.9567,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908293767603124e-05,
+      "loss": 3.9469,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8899907820092604e-05,
+      "loss": 3.945,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891521872582077e-05,
+      "loss": 3.9474,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883135925071557e-05,
+      "loss": 3.9503,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887474997756104e-05,
+      "loss": 3.9486,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886638040885425e-05,
+      "loss": 3.9464,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.885799446134373e-05,
+      "loss": 3.9417,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.884960851383321e-05,
+      "loss": 3.9458,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.884122256632269e-05,
+      "loss": 3.9507,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.88328529976159e-05,
+      "loss": 3.9389,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882446705010538e-05,
+      "loss": 3.9355,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881608110259486e-05,
+      "loss": 3.9435,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807711533888077e-05,
+      "loss": 3.9489,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879932558637755e-05,
+      "loss": 3.935,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879093963886703e-05,
+      "loss": 3.9515,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.878255369135651e-05,
+      "loss": 3.9402,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8774167743845994e-05,
+      "loss": 3.9518,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8765781796335474e-05,
+      "loss": 3.9471,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8757395848824954e-05,
+      "loss": 3.9372,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.008778095245361,
+      "eval_runtime": 294.7722,
+      "eval_samples_per_second": 1294.528,
+      "eval_steps_per_second": 40.455,
+      "step": 1297428
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8749009901314434e-05,
+      "loss": 3.9347,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8740640332607643e-05,
+      "loss": 3.9335,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8732254385097123e-05,
+      "loss": 3.9521,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8723868437586603e-05,
+      "loss": 3.9436,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715482490076086e-05,
+      "loss": 3.9478,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707096542565566e-05,
+      "loss": 3.9432,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8698710595055046e-05,
+      "loss": 3.9349,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8690341026348255e-05,
+      "loss": 3.9309,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8681971457641464e-05,
+      "loss": 3.9455,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8673585510130947e-05,
+      "loss": 3.9342,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8665199562620427e-05,
+      "loss": 3.9618,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8656813615109907e-05,
+      "loss": 3.9481,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8648444046403116e-05,
+      "loss": 3.9327,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640058098892596e-05,
+      "loss": 3.9355,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631672151382076e-05,
+      "loss": 3.9376,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8623286203871556e-05,
+      "loss": 3.9323,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.861490025636104e-05,
+      "loss": 3.9423,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.860651430885052e-05,
+      "loss": 3.9438,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.859812836134e-05,
+      "loss": 3.9316,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858974241382948e-05,
+      "loss": 3.959,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858137284512269e-05,
+      "loss": 3.9366,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.85730032764159e-05,
+      "loss": 3.9424,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.856461732890538e-05,
+      "loss": 3.9353,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.855623138139486e-05,
+      "loss": 3.9568,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.854784543388434e-05,
+      "loss": 3.9353,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.853945948637382e-05,
+      "loss": 3.9389,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.85310735388633e-05,
+      "loss": 3.9347,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.852270397015651e-05,
+      "loss": 3.9358,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.851431802264599e-05,
+      "loss": 3.9339,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8505932075135473e-05,
+      "loss": 3.9373,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497546127624953e-05,
+      "loss": 3.9381,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489160180114433e-05,
+      "loss": 3.9414,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8480774232603913e-05,
+      "loss": 3.943,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472388285093386e-05,
+      "loss": 3.9451,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464018716386602e-05,
+      "loss": 3.9428,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455632768876082e-05,
+      "loss": 3.9447,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.844724682136556e-05,
+      "loss": 3.9329,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.843886087385504e-05,
+      "loss": 3.9378,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.843047492634452e-05,
+      "loss": 3.9283,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8422088978834e-05,
+      "loss": 3.938,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.841370303132348e-05,
+      "loss": 3.9324,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.840531708381296e-05,
+      "loss": 3.9406,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8396931136302442e-05,
+      "loss": 3.9304,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8388545188791922e-05,
+      "loss": 3.9454,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.838017562008513e-05,
+      "loss": 3.9401,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.837178967257461e-05,
+      "loss": 3.9446,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.836340372506409e-05,
+      "loss": 3.9404,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.835501777755357e-05,
+      "loss": 3.9256,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8346648208846783e-05,
+      "loss": 3.9312,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8338278640139992e-05,
+      "loss": 3.9501,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8329892692629472e-05,
+      "loss": 3.9402,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8321506745118952e-05,
+      "loss": 3.9325,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8313120797608432e-05,
+      "loss": 3.929,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304734850097912e-05,
+      "loss": 3.9363,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296348902587396e-05,
+      "loss": 3.9187,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8287962955076875e-05,
+      "loss": 3.9492,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8279593386370085e-05,
+      "loss": 3.9172,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271207438859565e-05,
+      "loss": 3.9388,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8262821491349044e-05,
+      "loss": 3.9421,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8254435543838524e-05,
+      "loss": 3.9248,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8246049596328004e-05,
+      "loss": 3.9407,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8237663648817488e-05,
+      "loss": 3.9285,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8229277701306968e-05,
+      "loss": 3.9314,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8220908132600177e-05,
+      "loss": 3.9209,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212522185089657e-05,
+      "loss": 3.9443,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8204136237579137e-05,
+      "loss": 3.9274,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8195750290068617e-05,
+      "loss": 3.9344,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8187364342558097e-05,
+      "loss": 3.9186,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8178978395047573e-05,
+      "loss": 3.9256,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817060882634079e-05,
+      "loss": 3.9388,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.816222287883027e-05,
+      "loss": 3.9396,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8153836931319742e-05,
+      "loss": 3.9408,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8145450983809226e-05,
+      "loss": 3.9419,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8137065036298705e-05,
+      "loss": 3.9435,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128679088788185e-05,
+      "loss": 3.9349,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8120309520081395e-05,
+      "loss": 3.9449,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8111923572570874e-05,
+      "loss": 3.9328,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8103537625060354e-05,
+      "loss": 3.933,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8095151677549834e-05,
+      "loss": 3.9278,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8086765730039318e-05,
+      "loss": 3.938,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8078396161332527e-05,
+      "loss": 3.9375,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070010213822007e-05,
+      "loss": 3.9469,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8061624266311487e-05,
+      "loss": 3.9381,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8053238318800967e-05,
+      "loss": 3.9178,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8044852371290447e-05,
+      "loss": 3.924,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8036466423779927e-05,
+      "loss": 3.9344,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8028080476269407e-05,
+      "loss": 3.9331,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801971090756262e-05,
+      "loss": 3.9305,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.80113249600521e-05,
+      "loss": 3.9398,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.800293901254158e-05,
+      "loss": 3.9277,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.799455306503106e-05,
+      "loss": 3.9306,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.798616711752054e-05,
+      "loss": 3.9328,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.797779754881375e-05,
+      "loss": 3.9414,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796941160130323e-05,
+      "loss": 3.9415,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796102565379271e-05,
+      "loss": 3.9483,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.795263970628219e-05,
+      "loss": 3.9184,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.794425375877167e-05,
+      "loss": 3.9351,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.793586781126115e-05,
+      "loss": 3.9327,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.792748186375063e-05,
+      "loss": 3.9378,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919095916240108e-05,
+      "loss": 3.9469,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7910709968729588e-05,
+      "loss": 3.9237,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7902340400022803e-05,
+      "loss": 3.9327,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7893954452512283e-05,
+      "loss": 3.938,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885568505001757e-05,
+      "loss": 3.9397,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877182557491237e-05,
+      "loss": 3.9229,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786879660998072e-05,
+      "loss": 3.9328,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7860427041273936e-05,
+      "loss": 3.9259,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.785204109376341e-05,
+      "loss": 3.9277,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7843671525056625e-05,
+      "loss": 3.9269,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7835285577546105e-05,
+      "loss": 3.9291,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.782689963003558e-05,
+      "loss": 3.9268,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.781851368252506e-05,
+      "loss": 3.9361,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.781012773501454e-05,
+      "loss": 3.9288,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.780174178750402e-05,
+      "loss": 3.9319,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.77933558399935e-05,
+      "loss": 3.9455,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.778498627128671e-05,
+      "loss": 3.9411,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.777660032377619e-05,
+      "loss": 3.9353,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768214376265674e-05,
+      "loss": 3.9329,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7759828428755154e-05,
+      "loss": 3.9324,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7751442481244633e-05,
+      "loss": 3.9326,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7743056533734113e-05,
+      "loss": 3.923,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734686965027323e-05,
+      "loss": 3.932,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726301017516803e-05,
+      "loss": 3.9332,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7717915070006282e-05,
+      "loss": 3.9356,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7709529122495766e-05,
+      "loss": 3.9304,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7701143174985246e-05,
+      "loss": 3.9346,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7692757227474726e-05,
+      "loss": 3.9321,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7684371279964206e-05,
+      "loss": 3.9299,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7675985332453686e-05,
+      "loss": 3.9236,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667615763746895e-05,
+      "loss": 3.9471,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7659229816236375e-05,
+      "loss": 3.9406,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7650843868725858e-05,
+      "loss": 3.9357,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7642457921215338e-05,
+      "loss": 3.9329,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634088352508547e-05,
+      "loss": 3.938,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7625702404998027e-05,
+      "loss": 3.9374,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7617316457487507e-05,
+      "loss": 3.9406,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7608930509976987e-05,
+      "loss": 3.9341,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7600544562466467e-05,
+      "loss": 3.9347,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.759217499375968e-05,
+      "loss": 3.9336,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.758378904624916e-05,
+      "loss": 3.9302,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.757540309873864e-05,
+      "loss": 3.9285,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.756701715122812e-05,
+      "loss": 3.9342,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755864758252133e-05,
+      "loss": 3.9382,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755026163501081e-05,
+      "loss": 3.9259,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.754187568750029e-05,
+      "loss": 3.9419,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7533489739989765e-05,
+      "loss": 3.9324,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7525103792479245e-05,
+      "loss": 3.9388,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7516717844968725e-05,
+      "loss": 3.9375,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508331897458205e-05,
+      "loss": 3.9294,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.004150390625,
+      "eval_runtime": 296.2025,
+      "eval_samples_per_second": 1288.277,
+      "eval_steps_per_second": 40.26,
+      "step": 1373748
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7499945949947688e-05,
+      "loss": 3.9228,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491560002437168e-05,
+      "loss": 3.9232,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483174054926648e-05,
+      "loss": 3.9432,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7474788107416128e-05,
+      "loss": 3.933,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466402159905608e-05,
+      "loss": 3.94,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458016212395088e-05,
+      "loss": 3.9314,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7449630264884568e-05,
+      "loss": 3.9296,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7441244317374048e-05,
+      "loss": 3.9189,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7432858369863528e-05,
+      "loss": 3.9343,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7424472422353008e-05,
+      "loss": 3.9296,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7416086474842484e-05,
+      "loss": 3.951,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7407700527331964e-05,
+      "loss": 3.9388,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739933095862518e-05,
+      "loss": 3.9208,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7390945011114653e-05,
+      "loss": 3.9238,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7382559063604137e-05,
+      "loss": 3.9337,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374173116093617e-05,
+      "loss": 3.9137,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7365787168583096e-05,
+      "loss": 3.9346,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7357401221072576e-05,
+      "loss": 3.9309,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7349015273562056e-05,
+      "loss": 3.9249,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7340645704855265e-05,
+      "loss": 3.9527,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332259757344745e-05,
+      "loss": 3.9276,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.732387380983423e-05,
+      "loss": 3.9362,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.731548786232371e-05,
+      "loss": 3.9264,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.730710191481319e-05,
+      "loss": 3.9452,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7298732346106398e-05,
+      "loss": 3.921,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290346398595878e-05,
+      "loss": 3.9315,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7281960451085358e-05,
+      "loss": 3.9209,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7273574503574838e-05,
+      "loss": 3.9283,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.726520493486805e-05,
+      "loss": 3.928,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.725681898735753e-05,
+      "loss": 3.9238,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724843303984701e-05,
+      "loss": 3.932,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724004709233649e-05,
+      "loss": 3.935,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.723166114482597e-05,
+      "loss": 3.9288,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.722327519731545e-05,
+      "loss": 3.9417,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.721492200741239e-05,
+      "loss": 3.932,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.720653605990187e-05,
+      "loss": 3.9326,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.719815011239135e-05,
+      "loss": 3.926,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718976416488083e-05,
+      "loss": 3.9284,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718137821737031e-05,
+      "loss": 3.9202,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.717299226985979e-05,
+      "loss": 3.9283,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7164606322349275e-05,
+      "loss": 3.9223,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156220374838755e-05,
+      "loss": 3.9359,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7147834427328235e-05,
+      "loss": 3.9198,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139448479817715e-05,
+      "loss": 3.9338,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131062532307194e-05,
+      "loss": 3.9364,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122676584796668e-05,
+      "loss": 3.9324,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.711429063728615e-05,
+      "loss": 3.9326,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.710590468977563e-05,
+      "loss": 3.9148,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709751874226511e-05,
+      "loss": 3.9229,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.708914917355832e-05,
+      "loss": 3.9397,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7080779604851536e-05,
+      "loss": 3.929,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7072393657341016e-05,
+      "loss": 3.9266,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7064007709830492e-05,
+      "loss": 3.9166,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7055621762319972e-05,
+      "loss": 3.9273,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7047235814809452e-05,
+      "loss": 3.9086,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7038849867298932e-05,
+      "loss": 3.9409,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703048029859214e-05,
+      "loss": 3.9072,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.702209435108162e-05,
+      "loss": 3.9351,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7013708403571105e-05,
+      "loss": 3.9285,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7005322456060585e-05,
+      "loss": 3.9186,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6996936508550065e-05,
+      "loss": 3.9333,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6988550561039545e-05,
+      "loss": 3.9199,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6980164613529024e-05,
+      "loss": 3.919,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6971778666018504e-05,
+      "loss": 3.9138,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6963409097311714e-05,
+      "loss": 3.9353,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6955023149801193e-05,
+      "loss": 3.9184,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6946637202290677e-05,
+      "loss": 3.9239,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6938251254780157e-05,
+      "loss": 3.909,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6929881686073366e-05,
+      "loss": 3.9183,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921495738562846e-05,
+      "loss": 3.9297,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913109791052326e-05,
+      "loss": 3.9332,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6904723843541806e-05,
+      "loss": 3.9263,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6896354274835018e-05,
+      "loss": 3.9337,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6887968327324498e-05,
+      "loss": 3.9338,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6879582379813978e-05,
+      "loss": 3.9279,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.687122918991092e-05,
+      "loss": 3.9344,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.68628432424004e-05,
+      "loss": 3.9273,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.685445729488988e-05,
+      "loss": 3.9229,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.684607134737936e-05,
+      "loss": 3.9218,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.683768539986884e-05,
+      "loss": 3.9289,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682929945235832e-05,
+      "loss": 3.9304,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.68209135048478e-05,
+      "loss": 3.9397,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.681252755733728e-05,
+      "loss": 3.9262,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.680414160982676e-05,
+      "loss": 3.9076,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.679575566231624e-05,
+      "loss": 3.9138,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787369714805723e-05,
+      "loss": 3.9272,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779000146098932e-05,
+      "loss": 3.9224,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770614198588412e-05,
+      "loss": 3.9214,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6762228251077892e-05,
+      "loss": 3.9298,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.675384230356737e-05,
+      "loss": 3.9227,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.674545635605685e-05,
+      "loss": 3.916,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737070408546328e-05,
+      "loss": 3.9225,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728684461035808e-05,
+      "loss": 3.9368,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720298513525288e-05,
+      "loss": 3.9298,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6711928944818504e-05,
+      "loss": 3.9359,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703542997307977e-05,
+      "loss": 3.918,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.669515704979746e-05,
+      "loss": 3.9232,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.668677110228694e-05,
+      "loss": 3.9257,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.667840153358015e-05,
+      "loss": 3.9267,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.667001558606963e-05,
+      "loss": 3.9368,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.666162963855911e-05,
+      "loss": 3.9191,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.665324369104859e-05,
+      "loss": 3.9242,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.664485774353807e-05,
+      "loss": 3.9263,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6636488174831282e-05,
+      "loss": 3.9291,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6628102227320762e-05,
+      "loss": 3.9137,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6619716279810242e-05,
+      "loss": 3.9218,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6611330332299722e-05,
+      "loss": 3.9202,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.660296076359293e-05,
+      "loss": 3.9188,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6594574816082414e-05,
+      "loss": 3.9152,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6586205247375623e-05,
+      "loss": 3.9168,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6577819299865103e-05,
+      "loss": 3.9204,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6569433352354583e-05,
+      "loss": 3.9268,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6561047404844063e-05,
+      "loss": 3.9146,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6552661457333543e-05,
+      "loss": 3.9239,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6544291888626756e-05,
+      "loss": 3.9353,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6535905941116236e-05,
+      "loss": 3.9347,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6527519993605716e-05,
+      "loss": 3.9256,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519134046095195e-05,
+      "loss": 3.9262,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6510748098584675e-05,
+      "loss": 3.9236,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6502362151074155e-05,
+      "loss": 3.9217,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6493976203563635e-05,
+      "loss": 3.9165,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6485606634856848e-05,
+      "loss": 3.9197,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6477220687346328e-05,
+      "loss": 3.9211,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6468834739835808e-05,
+      "loss": 3.9313,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460448792325288e-05,
+      "loss": 3.9209,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452062844814768e-05,
+      "loss": 3.9251,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6443676897304248e-05,
+      "loss": 3.9235,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643530732859746e-05,
+      "loss": 3.9199,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642692138108694e-05,
+      "loss": 3.9147,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641853543357642e-05,
+      "loss": 3.9386,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64101494860659e-05,
+      "loss": 3.9345,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.640176353855538e-05,
+      "loss": 3.9293,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639337759104486e-05,
+      "loss": 3.9225,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6384991643534336e-05,
+      "loss": 3.9254,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.637662207482755e-05,
+      "loss": 3.9291,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368236127317032e-05,
+      "loss": 3.9394,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6359850179806512e-05,
+      "loss": 3.9239,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6351464232295985e-05,
+      "loss": 3.9212,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343078284785465e-05,
+      "loss": 3.9258,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334692337274945e-05,
+      "loss": 3.9235,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6326306389764425e-05,
+      "loss": 3.9201,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.631792044225391e-05,
+      "loss": 3.9246,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.630953449474339e-05,
+      "loss": 3.9292,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.630114854723287e-05,
+      "loss": 3.9165,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6292778978526078e-05,
+      "loss": 3.9326,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284393031015558e-05,
+      "loss": 3.9271,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276007083505038e-05,
+      "loss": 3.9241,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.626763751479825e-05,
+      "loss": 3.9292,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.625925156728773e-05,
+      "loss": 3.9158,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0000457763671875,
+      "eval_runtime": 309.8909,
+      "eval_samples_per_second": 1231.372,
+      "eval_steps_per_second": 38.481,
+      "step": 1450068
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.625086561977721e-05,
+      "loss": 3.9215,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.624247967226669e-05,
+      "loss": 3.9121,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.623409372475617e-05,
+      "loss": 3.9354,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.622570777724565e-05,
+      "loss": 3.9187,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.621732182973513e-05,
+      "loss": 3.9349,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.620893588222461e-05,
+      "loss": 3.9199,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6200566313517822e-05,
+      "loss": 3.9214,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6192180366007302e-05,
+      "loss": 3.9111,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6183794418496782e-05,
+      "loss": 3.927,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6175408470986262e-05,
+      "loss": 3.9145,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6167022523475742e-05,
+      "loss": 3.9394,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6158636575965222e-05,
+      "loss": 3.9279,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6150267007258434e-05,
+      "loss": 3.9112,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6141881059747914e-05,
+      "loss": 3.9164,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6133495112237394e-05,
+      "loss": 3.928,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6125109164726874e-05,
+      "loss": 3.909,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6116723217216347e-05,
+      "loss": 3.9236,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.610833726970583e-05,
+      "loss": 3.9212,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.609995132219531e-05,
+      "loss": 3.917,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6091581753488527e-05,
+      "loss": 3.9435,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6083195805978e-05,
+      "loss": 3.9215,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.607480985846748e-05,
+      "loss": 3.9213,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.606642391095696e-05,
+      "loss": 3.9167,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.605803796344644e-05,
+      "loss": 3.9365,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6049652015935923e-05,
+      "loss": 3.9144,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6041266068425403e-05,
+      "loss": 3.9193,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6032880120914883e-05,
+      "loss": 3.9149,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6024494173404363e-05,
+      "loss": 3.9198,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6016124604697572e-05,
+      "loss": 3.9184,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6007755035990784e-05,
+      "loss": 3.9177,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5999369088480264e-05,
+      "loss": 3.9231,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5990983140969744e-05,
+      "loss": 3.9231,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5982597193459224e-05,
+      "loss": 3.9235,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5974211245948704e-05,
+      "loss": 3.9305,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5965841677241913e-05,
+      "loss": 3.9224,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957455729731393e-05,
+      "loss": 3.9265,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5949069782220877e-05,
+      "loss": 3.9112,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5940683834710357e-05,
+      "loss": 3.9234,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5932314266003566e-05,
+      "loss": 3.9104,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5923928318493046e-05,
+      "loss": 3.9168,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5915542370982526e-05,
+      "loss": 3.919,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5907156423472006e-05,
+      "loss": 3.928,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5898770475961486e-05,
+      "loss": 3.9089,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5890384528450966e-05,
+      "loss": 3.9218,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.588199858094045e-05,
+      "loss": 3.9298,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.587361263342993e-05,
+      "loss": 3.9232,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.586522668591941e-05,
+      "loss": 3.928,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5856873496016347e-05,
+      "loss": 3.9064,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5848487548505827e-05,
+      "loss": 3.9124,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.584010160099531e-05,
+      "loss": 3.9271,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.583171565348479e-05,
+      "loss": 3.9203,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.582332970597427e-05,
+      "loss": 3.9179,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.581494375846375e-05,
+      "loss": 3.9103,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.580655781095323e-05,
+      "loss": 3.9219,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.579817186344271e-05,
+      "loss": 3.8997,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.578980229473592e-05,
+      "loss": 3.9268,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781416347225403e-05,
+      "loss": 3.905,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5773030399714882e-05,
+      "loss": 3.9194,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.576466083100809e-05,
+      "loss": 3.9241,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.575627488349757e-05,
+      "loss": 3.9097,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.574788893598705e-05,
+      "loss": 3.922,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.573950298847653e-05,
+      "loss": 3.9136,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731117040966008e-05,
+      "loss": 3.9093,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5722731093455488e-05,
+      "loss": 3.9063,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5714345145944968e-05,
+      "loss": 3.9232,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5705959198434448e-05,
+      "loss": 3.9119,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5697589629727657e-05,
+      "loss": 3.9156,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.568920368221714e-05,
+      "loss": 3.9031,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.568081773470662e-05,
+      "loss": 3.9071,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.56724317871961e-05,
+      "loss": 3.9187,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.566404583968558e-05,
+      "loss": 3.9259,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.565567627097879e-05,
+      "loss": 3.9136,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.564729032346827e-05,
+      "loss": 3.9277,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5638920754761482e-05,
+      "loss": 3.9258,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5630534807250962e-05,
+      "loss": 3.9164,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.562214885974044e-05,
+      "loss": 3.9218,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.561376291222992e-05,
+      "loss": 3.9185,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.560539334352313e-05,
+      "loss": 3.914,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.559700739601261e-05,
+      "loss": 3.9136,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5588621448502094e-05,
+      "loss": 3.9185,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5580235500991574e-05,
+      "loss": 3.923,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5571849553481054e-05,
+      "loss": 3.9282,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5563463605970534e-05,
+      "loss": 3.919,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5555094037263743e-05,
+      "loss": 3.8986,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5546708089753223e-05,
+      "loss": 3.9082,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5538322142242703e-05,
+      "loss": 3.9129,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5529936194732186e-05,
+      "loss": 3.9199,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5521550247221666e-05,
+      "loss": 3.9096,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5513164299711146e-05,
+      "loss": 3.9201,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5504778352200626e-05,
+      "loss": 3.9138,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5496408783493835e-05,
+      "loss": 3.9077,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5488022835983315e-05,
+      "loss": 3.9167,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5479636888472795e-05,
+      "loss": 3.9216,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.547125094096228e-05,
+      "loss": 3.9193,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.546286499345176e-05,
+      "loss": 3.9308,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.545447904594124e-05,
+      "loss": 3.9115,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.544609309843072e-05,
+      "loss": 3.9124,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.543770715092019e-05,
+      "loss": 3.9138,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.542932120340967e-05,
+      "loss": 3.9201,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5420951634702887e-05,
+      "loss": 3.9263,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.54125820659961e-05,
+      "loss": 3.911,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.540419611848558e-05,
+      "loss": 3.9151,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.539581017097506e-05,
+      "loss": 3.9177,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.538742422346454e-05,
+      "loss": 3.9215,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.537903827595402e-05,
+      "loss": 3.9042,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5370652328443496e-05,
+      "loss": 3.915,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5362266380932976e-05,
+      "loss": 3.9142,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5353880433422456e-05,
+      "loss": 3.9091,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5345510864715665e-05,
+      "loss": 3.9046,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5337124917205145e-05,
+      "loss": 3.9106,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5328738969694625e-05,
+      "loss": 3.915,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.532035302218411e-05,
+      "loss": 3.9137,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5311983453477318e-05,
+      "loss": 3.9099,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5303597505966798e-05,
+      "loss": 3.9129,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5295211558456277e-05,
+      "loss": 3.9288,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5286825610945757e-05,
+      "loss": 3.9243,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5278439663435237e-05,
+      "loss": 3.9172,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.527007009472845e-05,
+      "loss": 3.9145,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.526168414721793e-05,
+      "loss": 3.9135,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.525329819970741e-05,
+      "loss": 3.9149,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.524491225219689e-05,
+      "loss": 3.9069,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.523652630468637e-05,
+      "loss": 3.9098,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.522815673597958e-05,
+      "loss": 3.9108,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.521977078846906e-05,
+      "loss": 3.9242,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5211384840958542e-05,
+      "loss": 3.9124,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5202998893448022e-05,
+      "loss": 3.92,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.519462932474123e-05,
+      "loss": 3.9121,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.518624337723071e-05,
+      "loss": 3.9098,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.517787380852392e-05,
+      "loss": 3.9083,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5169487861013404e-05,
+      "loss": 3.9279,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5161101913502883e-05,
+      "loss": 3.9302,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5152715965992363e-05,
+      "loss": 3.9167,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5144346397285573e-05,
+      "loss": 3.915,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5135960449775052e-05,
+      "loss": 3.9175,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5127574502264532e-05,
+      "loss": 3.9164,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5119188554754012e-05,
+      "loss": 3.9307,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5110802607243496e-05,
+      "loss": 3.9159,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5102416659732976e-05,
+      "loss": 3.9142,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5094030712222456e-05,
+      "loss": 3.9126,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5085644764711936e-05,
+      "loss": 3.9179,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5077258817201416e-05,
+      "loss": 3.9098,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5068889248494625e-05,
+      "loss": 3.9131,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5060503300984105e-05,
+      "loss": 3.9211,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5052133732277317e-05,
+      "loss": 3.9109,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5043747784766797e-05,
+      "loss": 3.9198,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5035361837256277e-05,
+      "loss": 3.9242,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5026975889745757e-05,
+      "loss": 3.9159,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5018589942235237e-05,
+      "loss": 3.9195,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5010203994724717e-05,
+      "loss": 3.9084,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9958603382110596,
+      "eval_runtime": 314.3894,
+      "eval_samples_per_second": 1213.753,
+      "eval_steps_per_second": 37.931,
+      "step": 1526388
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5001818047214197e-05,
+      "loss": 3.9052,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4993432099703677e-05,
+      "loss": 3.9055,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4985046152193157e-05,
+      "loss": 3.9269,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4976660204682637e-05,
+      "loss": 3.9146,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4968274257172117e-05,
+      "loss": 3.9274,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4959888309661593e-05,
+      "loss": 3.9083,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4951502362151073e-05,
+      "loss": 3.9173,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4943116414640557e-05,
+      "loss": 3.8989,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4934730467130036e-05,
+      "loss": 3.9192,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4926344519619516e-05,
+      "loss": 3.9034,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4917958572108996e-05,
+      "loss": 3.9288,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4909572624598476e-05,
+      "loss": 3.9249,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4901203055891685e-05,
+      "loss": 3.8985,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4892817108381165e-05,
+      "loss": 3.9099,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.488443116087065e-05,
+      "loss": 3.9199,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4876045213360125e-05,
+      "loss": 3.9032,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4867659265849605e-05,
+      "loss": 3.9105,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4859273318339085e-05,
+      "loss": 3.9147,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4850887370828565e-05,
+      "loss": 3.9114,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4842501423318045e-05,
+      "loss": 3.932,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4834115475807525e-05,
+      "loss": 3.9127,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4825729528297005e-05,
+      "loss": 3.9126,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4817359959590218e-05,
+      "loss": 3.9152,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4808974012079697e-05,
+      "loss": 3.9254,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4800588064569177e-05,
+      "loss": 3.9082,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4792202117058654e-05,
+      "loss": 3.9164,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4783816169548134e-05,
+      "loss": 3.9019,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4775430222037614e-05,
+      "loss": 3.91,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.476706065333083e-05,
+      "loss": 3.9068,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4758674705820306e-05,
+      "loss": 3.9124,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4750288758309786e-05,
+      "loss": 3.9089,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4741902810799266e-05,
+      "loss": 3.92,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4733516863288746e-05,
+      "loss": 3.9166,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4725130915778226e-05,
+      "loss": 3.9229,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.471676134707144e-05,
+      "loss": 3.9092,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.470837539956092e-05,
+      "loss": 3.9187,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.46999894520504e-05,
+      "loss": 3.9053,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.469160350453988e-05,
+      "loss": 3.9157,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.468321755702936e-05,
+      "loss": 3.903,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4674831609518835e-05,
+      "loss": 3.9111,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.466644566200832e-05,
+      "loss": 3.9074,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.46580597144978e-05,
+      "loss": 3.9204,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4649673766987278e-05,
+      "loss": 3.9066,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4641304198280487e-05,
+      "loss": 3.9082,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.46329346295737e-05,
+      "loss": 3.9238,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.462454868206318e-05,
+      "loss": 3.9142,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.461616273455266e-05,
+      "loss": 3.9189,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.460777678704214e-05,
+      "loss": 3.9039,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.459939083953162e-05,
+      "loss": 3.899,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.45910048920211e-05,
+      "loss": 3.9187,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.458261894451058e-05,
+      "loss": 3.9137,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.457423299700006e-05,
+      "loss": 3.9085,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.456586342829327e-05,
+      "loss": 3.9059,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557477480782752e-05,
+      "loss": 3.914,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549091533272232e-05,
+      "loss": 3.8957,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4540705585761712e-05,
+      "loss": 3.9154,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.453233601705492e-05,
+      "loss": 3.8965,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.45239500695444e-05,
+      "loss": 3.908,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.451556412203388e-05,
+      "loss": 3.9179,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.450717817452336e-05,
+      "loss": 3.9024,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4498808605816573e-05,
+      "loss": 3.9111,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4490422658306053e-05,
+      "loss": 3.9085,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4482036710795533e-05,
+      "loss": 3.899,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4473650763285013e-05,
+      "loss": 3.9049,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4465281194578222e-05,
+      "loss": 3.9121,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4456895247067706e-05,
+      "loss": 3.9045,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4448509299557186e-05,
+      "loss": 3.907,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4440123352046666e-05,
+      "loss": 3.8998,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4431737404536142e-05,
+      "loss": 3.8989,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4423351457025622e-05,
+      "loss": 3.9097,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4414965509515102e-05,
+      "loss": 3.9187,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4406579562004582e-05,
+      "loss": 3.9053,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4398193614494065e-05,
+      "loss": 3.9214,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4389807666983545e-05,
+      "loss": 3.9163,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4381438098276754e-05,
+      "loss": 3.9092,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373068529569964e-05,
+      "loss": 3.9139,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4364682582059443e-05,
+      "loss": 3.9189,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4356296634548927e-05,
+      "loss": 3.9067,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4347910687038407e-05,
+      "loss": 3.9071,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4339524739527887e-05,
+      "loss": 3.9119,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4331138792017367e-05,
+      "loss": 3.9127,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4322752844506847e-05,
+      "loss": 3.9156,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4314366896996323e-05,
+      "loss": 3.9141,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4305997328289536e-05,
+      "loss": 3.8947,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4297611380779016e-05,
+      "loss": 3.8938,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4289241812072228e-05,
+      "loss": 3.909,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4280855864561708e-05,
+      "loss": 3.9121,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4272486295854917e-05,
+      "loss": 3.9048,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4264100348344397e-05,
+      "loss": 3.9113,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.425571440083388e-05,
+      "loss": 3.908,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.424732845332336e-05,
+      "loss": 3.8987,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.423894250581284e-05,
+      "loss": 3.908,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.423055655830232e-05,
+      "loss": 3.9098,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4222170610791797e-05,
+      "loss": 3.916,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4213784663281277e-05,
+      "loss": 3.9152,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205398715770757e-05,
+      "loss": 3.9138,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4197012768260237e-05,
+      "loss": 3.8995,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.418862682074972e-05,
+      "loss": 3.9115,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.41802408732392e-05,
+      "loss": 3.9103,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.417187130453241e-05,
+      "loss": 3.919,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.416348535702189e-05,
+      "loss": 3.9068,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4155115788315098e-05,
+      "loss": 3.903,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.414672984080458e-05,
+      "loss": 3.9117,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.413834389329406e-05,
+      "loss": 3.9128,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.412995794578354e-05,
+      "loss": 3.9035,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.412157199827302e-05,
+      "loss": 3.906,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4113218808369963e-05,
+      "loss": 3.9062,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4104832860859443e-05,
+      "loss": 3.8995,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4096446913348923e-05,
+      "loss": 3.8973,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4088060965838403e-05,
+      "loss": 3.9071,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4079675018327883e-05,
+      "loss": 3.9049,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4071289070817363e-05,
+      "loss": 3.9057,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4062903123306843e-05,
+      "loss": 3.9033,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4054517175796323e-05,
+      "loss": 3.9052,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.40461312282858e-05,
+      "loss": 3.9211,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4037745280775283e-05,
+      "loss": 3.9211,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4029359333264763e-05,
+      "loss": 3.9116,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4020989764557975e-05,
+      "loss": 3.909,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4012620195851184e-05,
+      "loss": 3.9048,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4004234248340664e-05,
+      "loss": 3.9065,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3995848300830144e-05,
+      "loss": 3.9021,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3987462353319624e-05,
+      "loss": 3.9061,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3979076405809104e-05,
+      "loss": 3.9049,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3970690458298584e-05,
+      "loss": 3.9126,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3962304510788064e-05,
+      "loss": 3.9044,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3953918563277544e-05,
+      "loss": 3.9123,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3945532615767024e-05,
+      "loss": 3.9039,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3937146668256504e-05,
+      "loss": 3.9038,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3928760720745984e-05,
+      "loss": 3.9008,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3920391152039196e-05,
+      "loss": 3.9193,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3912005204528676e-05,
+      "loss": 3.9229,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3903619257018156e-05,
+      "loss": 3.9143,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3895233309507633e-05,
+      "loss": 3.9025,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3886847361997113e-05,
+      "loss": 3.913,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3878461414486593e-05,
+      "loss": 3.9117,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3870075466976076e-05,
+      "loss": 3.9211,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3861689519465556e-05,
+      "loss": 3.9072,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3853319950758765e-05,
+      "loss": 3.9057,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3844934003248245e-05,
+      "loss": 3.9084,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3836548055737725e-05,
+      "loss": 3.9116,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3828162108227205e-05,
+      "loss": 3.9016,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3819776160716685e-05,
+      "loss": 3.9041,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3811406592009897e-05,
+      "loss": 3.9135,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3803020644499377e-05,
+      "loss": 3.906,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3794634696988857e-05,
+      "loss": 3.9124,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3786248749478337e-05,
+      "loss": 3.917,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3777879180771546e-05,
+      "loss": 3.9082,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.376949323326103e-05,
+      "loss": 3.9071,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.376110728575051e-05,
+      "loss": 3.9068,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9933674335479736,
+      "eval_runtime": 299.3949,
+      "eval_samples_per_second": 1274.541,
+      "eval_steps_per_second": 39.83,
+      "step": 1602708
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3752721338239986e-05,
+      "loss": 3.8996,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3744335390729466e-05,
+      "loss": 3.8963,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3735949443218946e-05,
+      "loss": 3.9181,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3727563495708426e-05,
+      "loss": 3.9084,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3719177548197906e-05,
+      "loss": 3.9167,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3710791600687386e-05,
+      "loss": 3.9065,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.370240565317687e-05,
+      "loss": 3.9075,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3694019705666346e-05,
+      "loss": 3.8923,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3685633758155826e-05,
+      "loss": 3.9095,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3677247810645306e-05,
+      "loss": 3.9013,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3668861863134786e-05,
+      "loss": 3.9173,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3660475915624266e-05,
+      "loss": 3.917,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3652106346917478e-05,
+      "loss": 3.8949,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3643720399406958e-05,
+      "loss": 3.9024,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3635334451896438e-05,
+      "loss": 3.9121,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3626948504385918e-05,
+      "loss": 3.8959,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3618562556875398e-05,
+      "loss": 3.9044,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3610176609364875e-05,
+      "loss": 3.9064,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3601790661854355e-05,
+      "loss": 3.9052,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3593404714343838e-05,
+      "loss": 3.9221,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3585018766833318e-05,
+      "loss": 3.9081,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3576632819322798e-05,
+      "loss": 3.9077,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3568246871812278e-05,
+      "loss": 3.911,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3559860924301758e-05,
+      "loss": 3.9126,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3551491355594967e-05,
+      "loss": 3.9024,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3543105408084447e-05,
+      "loss": 3.9078,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.353471946057393e-05,
+      "loss": 3.8985,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.352634989186714e-05,
+      "loss": 3.9025,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.351796394435662e-05,
+      "loss": 3.9011,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.35095779968461e-05,
+      "loss": 3.9065,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.350119204933558e-05,
+      "loss": 3.9039,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3492806101825056e-05,
+      "loss": 3.9113,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.348442015431454e-05,
+      "loss": 3.9113,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.347603420680402e-05,
+      "loss": 3.9134,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.346766463809723e-05,
+      "loss": 3.9054,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3459278690586708e-05,
+      "loss": 3.9106,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3450892743076188e-05,
+      "loss": 3.9012,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3442506795565668e-05,
+      "loss": 3.9048,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3434120848055148e-05,
+      "loss": 3.896,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.342575127934836e-05,
+      "loss": 3.9001,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.341736533183784e-05,
+      "loss": 3.9006,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.340897938432732e-05,
+      "loss": 3.9129,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.34005934368168e-05,
+      "loss": 3.8971,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.339220748930628e-05,
+      "loss": 3.9046,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.338382154179576e-05,
+      "loss": 3.918,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.337543559428524e-05,
+      "loss": 3.9078,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.336704964677472e-05,
+      "loss": 3.9111,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3358680078067932e-05,
+      "loss": 3.8998,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.335031050936114e-05,
+      "loss": 3.887,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.334192456185062e-05,
+      "loss": 3.9169,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.33335386143401e-05,
+      "loss": 3.9039,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3325152666829585e-05,
+      "loss": 3.898,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.331676671931906e-05,
+      "loss": 3.9016,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.330838077180854e-05,
+      "loss": 3.9091,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.329999482429802e-05,
+      "loss": 3.8833,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3291625255591234e-05,
+      "loss": 3.9098,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3283239308080714e-05,
+      "loss": 3.8895,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3274853360570194e-05,
+      "loss": 3.9045,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3266467413059674e-05,
+      "loss": 3.9081,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3258081465549154e-05,
+      "loss": 3.9033,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3249695518038634e-05,
+      "loss": 3.9001,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3241309570528114e-05,
+      "loss": 3.9052,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3232923623017593e-05,
+      "loss": 3.8881,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3224554054310803e-05,
+      "loss": 3.8992,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3216168106800286e-05,
+      "loss": 3.9051,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3207782159289766e-05,
+      "loss": 3.9029,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3199396211779242e-05,
+      "loss": 3.8974,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3191026643072455e-05,
+      "loss": 3.8926,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3182640695561935e-05,
+      "loss": 3.8905,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3174254748051415e-05,
+      "loss": 3.9008,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3165868800540895e-05,
+      "loss": 3.9094,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3157482853030375e-05,
+      "loss": 3.8977,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3149113284323587e-05,
+      "loss": 3.9149,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3140727336813067e-05,
+      "loss": 3.9115,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3132341389302544e-05,
+      "loss": 3.9033,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3123955441792024e-05,
+      "loss": 3.9027,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.311558587308524e-05,
+      "loss": 3.9096,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3107199925574716e-05,
+      "loss": 3.9013,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3098813978064196e-05,
+      "loss": 3.8977,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3090428030553676e-05,
+      "loss": 3.9072,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3082042083043156e-05,
+      "loss": 3.9046,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3073656135532636e-05,
+      "loss": 3.9103,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.306528656682585e-05,
+      "loss": 3.9069,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.305690061931533e-05,
+      "loss": 3.8902,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.304851467180481e-05,
+      "loss": 3.8905,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3040145103098017e-05,
+      "loss": 3.8952,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.303177553439123e-05,
+      "loss": 3.9044,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.302338958688071e-05,
+      "loss": 3.9005,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3015020018173922e-05,
+      "loss": 3.9043,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3006634070663402e-05,
+      "loss": 3.9009,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2998248123152882e-05,
+      "loss": 3.8949,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2989862175642362e-05,
+      "loss": 3.9027,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.298147622813184e-05,
+      "loss": 3.8998,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2973090280621322e-05,
+      "loss": 3.9091,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2964704333110802e-05,
+      "loss": 3.911,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2956318385600282e-05,
+      "loss": 3.9073,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.294794881689349e-05,
+      "loss": 3.8946,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.293956286938297e-05,
+      "loss": 3.9023,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.293117692187245e-05,
+      "loss": 3.9047,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.292279097436193e-05,
+      "loss": 3.9125,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.291440502685141e-05,
+      "loss": 3.902,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2906019079340894e-05,
+      "loss": 3.894,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.289763313183037e-05,
+      "loss": 3.9118,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2889263563123583e-05,
+      "loss": 3.9007,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2880877615613063e-05,
+      "loss": 3.8982,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2872491668102543e-05,
+      "loss": 3.9019,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2864105720592023e-05,
+      "loss": 3.8996,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2855736151885236e-05,
+      "loss": 3.8906,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2847350204374716e-05,
+      "loss": 3.886,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2838964256864192e-05,
+      "loss": 3.9025,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2830578309353672e-05,
+      "loss": 3.902,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2822192361843152e-05,
+      "loss": 3.897,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2813806414332632e-05,
+      "loss": 3.895,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2805420466822115e-05,
+      "loss": 3.8951,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2797034519311595e-05,
+      "loss": 3.9139,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2788664950604805e-05,
+      "loss": 3.9153,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2780279003094284e-05,
+      "loss": 3.9092,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2771893055583764e-05,
+      "loss": 3.9039,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2763507108073244e-05,
+      "loss": 3.8955,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2755137539366457e-05,
+      "loss": 3.8985,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2746751591855937e-05,
+      "loss": 3.8953,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2738365644345417e-05,
+      "loss": 3.9005,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2729979696834897e-05,
+      "loss": 3.8961,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2721593749324373e-05,
+      "loss": 3.9041,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2713207801813853e-05,
+      "loss": 3.9011,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2704821854303333e-05,
+      "loss": 3.9017,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2696435906792817e-05,
+      "loss": 3.9009,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2688066338086026e-05,
+      "loss": 3.8984,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2679680390575506e-05,
+      "loss": 3.8976,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2671294443064986e-05,
+      "loss": 3.9097,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2662908495554466e-05,
+      "loss": 3.9127,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2654538926847678e-05,
+      "loss": 3.911,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2646152979337158e-05,
+      "loss": 3.8955,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2637767031826638e-05,
+      "loss": 3.9093,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2629381084316118e-05,
+      "loss": 3.9027,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2621011515609327e-05,
+      "loss": 3.9127,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2612625568098807e-05,
+      "loss": 3.902,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2604239620588287e-05,
+      "loss": 3.9015,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.259585367307777e-05,
+      "loss": 3.9004,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.258746772556725e-05,
+      "loss": 3.9065,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257909815686046e-05,
+      "loss": 3.9001,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257071220934994e-05,
+      "loss": 3.8998,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.256232626183942e-05,
+      "loss": 3.8995,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.25539403143289e-05,
+      "loss": 3.9046,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.254557074562211e-05,
+      "loss": 3.9049,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.253718479811159e-05,
+      "loss": 3.904,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.252879885060107e-05,
+      "loss": 3.908,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.252041290309055e-05,
+      "loss": 3.8987,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.251204333438376e-05,
+      "loss": 3.9023,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9909539222717285,
+      "eval_runtime": 295.6924,
+      "eval_samples_per_second": 1290.5,
+      "eval_steps_per_second": 40.329,
+      "step": 1679028
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.250365738687324e-05,
+      "loss": 3.8958,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2495271439362724e-05,
+      "loss": 3.8898,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2486885491852204e-05,
+      "loss": 3.9085,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247849954434168e-05,
+      "loss": 3.9039,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247011359683116e-05,
+      "loss": 3.9123,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.246172764932064e-05,
+      "loss": 3.8986,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.245334170181012e-05,
+      "loss": 3.8933,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.24449557542996e-05,
+      "loss": 3.895,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.243656980678908e-05,
+      "loss": 3.8996,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.242818385927856e-05,
+      "loss": 3.8968,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.241979791176804e-05,
+      "loss": 3.9108,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.241141196425752e-05,
+      "loss": 3.9131,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2403042395550733e-05,
+      "loss": 3.889,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.239465644804021e-05,
+      "loss": 3.8959,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.238627050052969e-05,
+      "loss": 3.9088,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2377884553019172e-05,
+      "loss": 3.886,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2369498605508652e-05,
+      "loss": 3.8988,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2361112657998132e-05,
+      "loss": 3.8988,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2352726710487612e-05,
+      "loss": 3.8971,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2344340762977092e-05,
+      "loss": 3.9169,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233595481546657e-05,
+      "loss": 3.9026,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.232758524675978e-05,
+      "loss": 3.9037,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2319199299249265e-05,
+      "loss": 3.9013,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231081335173874e-05,
+      "loss": 3.9057,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.230242740422822e-05,
+      "loss": 3.8991,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.22940414567177e-05,
+      "loss": 3.899,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2285671888010914e-05,
+      "loss": 3.8943,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2277285940500394e-05,
+      "loss": 3.891,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2268899992989874e-05,
+      "loss": 3.8955,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2260514045479353e-05,
+      "loss": 3.8984,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2252144476772566e-05,
+      "loss": 3.9008,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2243758529262043e-05,
+      "loss": 3.901,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2235372581751522e-05,
+      "loss": 3.9049,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2226986634241002e-05,
+      "loss": 3.9086,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2218617065534215e-05,
+      "loss": 3.898,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2210231118023695e-05,
+      "loss": 3.9015,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2201845170513175e-05,
+      "loss": 3.8987,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2193459223002655e-05,
+      "loss": 3.8978,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2185089654295864e-05,
+      "loss": 3.8888,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2176703706785347e-05,
+      "loss": 3.8937,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2168317759274827e-05,
+      "loss": 3.8951,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2159931811764307e-05,
+      "loss": 3.9039,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2151545864253787e-05,
+      "loss": 3.894,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2143159916743267e-05,
+      "loss": 3.8958,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2134773969232747e-05,
+      "loss": 3.9166,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2126388021722224e-05,
+      "loss": 3.8982,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2118002074211704e-05,
+      "loss": 3.901,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.210963250550492e-05,
+      "loss": 3.8926,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.210126293679813e-05,
+      "loss": 3.8819,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2092893368091338e-05,
+      "loss": 3.9102,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2084507420580818e-05,
+      "loss": 3.8969,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2076121473070298e-05,
+      "loss": 3.8918,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.206773552555978e-05,
+      "loss": 3.8953,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205934957804926e-05,
+      "loss": 3.9045,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205096363053874e-05,
+      "loss": 3.878,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.204257768302822e-05,
+      "loss": 3.8944,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2034191735517697e-05,
+      "loss": 3.8895,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2025805788007177e-05,
+      "loss": 3.8885,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2017419840496657e-05,
+      "loss": 3.9069,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.200905027178987e-05,
+      "loss": 3.8935,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.200066432427935e-05,
+      "loss": 3.8911,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.199227837676883e-05,
+      "loss": 3.9017,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.198389242925831e-05,
+      "loss": 3.8853,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.197550648174779e-05,
+      "loss": 3.8881,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.196712053423727e-05,
+      "loss": 3.8975,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.195873458672675e-05,
+      "loss": 3.8947,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.195034863921623e-05,
+      "loss": 3.8959,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.194196269170571e-05,
+      "loss": 3.8878,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1933593122998922e-05,
+      "loss": 3.8794,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1925207175488402e-05,
+      "loss": 3.8939,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.191682122797788e-05,
+      "loss": 3.906,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1908435280467358e-05,
+      "loss": 3.8919,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.190004933295684e-05,
+      "loss": 3.9084,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.189167976425005e-05,
+      "loss": 3.9053,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.188329381673953e-05,
+      "loss": 3.8979,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1874924248032743e-05,
+      "loss": 3.891,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1866538300522223e-05,
+      "loss": 3.9079,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1858152353011703e-05,
+      "loss": 3.8939,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1849766405501183e-05,
+      "loss": 3.8903,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1841380457990663e-05,
+      "loss": 3.9015,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1832994510480143e-05,
+      "loss": 3.8971,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1824608562969623e-05,
+      "loss": 3.9059,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1816222615459103e-05,
+      "loss": 3.8987,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1807853046752312e-05,
+      "loss": 3.884,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1799467099241795e-05,
+      "loss": 3.8841,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1791081151731275e-05,
+      "loss": 3.8891,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1782695204220755e-05,
+      "loss": 3.898,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1774309256710232e-05,
+      "loss": 3.8937,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1765939688003444e-05,
+      "loss": 3.8923,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1757570119296657e-05,
+      "loss": 3.898,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1749184171786137e-05,
+      "loss": 3.8839,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1740798224275617e-05,
+      "loss": 3.8956,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1732412276765097e-05,
+      "loss": 3.8928,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1724026329254577e-05,
+      "loss": 3.9071,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1715640381744053e-05,
+      "loss": 3.9056,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1707254434233533e-05,
+      "loss": 3.8999,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1698868486723013e-05,
+      "loss": 3.8891,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169049891801623e-05,
+      "loss": 3.8971,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1682112970505705e-05,
+      "loss": 3.8932,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1673727022995185e-05,
+      "loss": 3.9083,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1665357454288398e-05,
+      "loss": 3.8955,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1656971506777878e-05,
+      "loss": 3.8898,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1648585559267358e-05,
+      "loss": 3.8999,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1640199611756838e-05,
+      "loss": 3.8972,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1631813664246318e-05,
+      "loss": 3.8908,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1623427716735798e-05,
+      "loss": 3.8971,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1615041769225278e-05,
+      "loss": 3.8862,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1606655821714758e-05,
+      "loss": 3.8869,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1598286253007967e-05,
+      "loss": 3.8794,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.158990030549745e-05,
+      "loss": 3.8981,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.158151435798693e-05,
+      "loss": 3.8934,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.157312841047641e-05,
+      "loss": 3.8918,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.156475884176962e-05,
+      "loss": 3.89,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.15563728942591e-05,
+      "loss": 3.8877,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.154798694674858e-05,
+      "loss": 3.9034,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.153961737804179e-05,
+      "loss": 3.9098,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.153123143053127e-05,
+      "loss": 3.9046,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.152284548302075e-05,
+      "loss": 3.8944,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.151445953551023e-05,
+      "loss": 3.8911,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.150608996680344e-05,
+      "loss": 3.8936,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1497720398096653e-05,
+      "loss": 3.8845,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1489334450586133e-05,
+      "loss": 3.8951,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1480948503075613e-05,
+      "loss": 3.8915,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1472562555565093e-05,
+      "loss": 3.8947,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1464176608054573e-05,
+      "loss": 3.8984,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1455790660544053e-05,
+      "loss": 3.8911,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1447404713033533e-05,
+      "loss": 3.8979,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1439018765523013e-05,
+      "loss": 3.8889,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1430649196816225e-05,
+      "loss": 3.8934,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1422263249305705e-05,
+      "loss": 3.8971,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1413893680598914e-05,
+      "loss": 3.9055,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1405507733088394e-05,
+      "loss": 3.9021,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1397121785577874e-05,
+      "loss": 3.8915,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1388735838067357e-05,
+      "loss": 3.9022,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1380349890556834e-05,
+      "loss": 3.8987,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1371963943046314e-05,
+      "loss": 3.9032,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1363577995535794e-05,
+      "loss": 3.8975,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1355192048025274e-05,
+      "loss": 3.8982,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1346806100514754e-05,
+      "loss": 3.8913,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1338420153004234e-05,
+      "loss": 3.8985,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1330050584297446e-05,
+      "loss": 3.8889,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1321664636786926e-05,
+      "loss": 3.8939,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1313278689276406e-05,
+      "loss": 3.8963,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1304892741765886e-05,
+      "loss": 3.8972,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1296523173059095e-05,
+      "loss": 3.8975,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1288137225548575e-05,
+      "loss": 3.8962,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.127975127803806e-05,
+      "loss": 3.9032,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.127136533052754e-05,
+      "loss": 3.887,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1262995761820748e-05,
+      "loss": 3.9014,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.98861026763916,
+      "eval_runtime": 289.2882,
+      "eval_samples_per_second": 1319.069,
+      "eval_steps_per_second": 41.222,
+      "step": 1755348
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.125464257191769e-05,
+      "loss": 3.8931,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1246256624407173e-05,
+      "loss": 3.8811,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1237870676896652e-05,
+      "loss": 3.901,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.122948472938613e-05,
+      "loss": 3.8972,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.122109878187561e-05,
+      "loss": 3.9039,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.121271283436509e-05,
+      "loss": 3.892,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.120432688685457e-05,
+      "loss": 3.8888,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.119594093934405e-05,
+      "loss": 3.8895,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.118755499183353e-05,
+      "loss": 3.8936,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1179169044323012e-05,
+      "loss": 3.8882,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.117078309681249e-05,
+      "loss": 3.9068,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.116239714930197e-05,
+      "loss": 3.9031,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.115402758059518e-05,
+      "loss": 3.8881,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1145641633084658e-05,
+      "loss": 3.8876,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1137255685574138e-05,
+      "loss": 3.9001,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.112886973806362e-05,
+      "loss": 3.8818,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.11204837905531e-05,
+      "loss": 3.8935,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.111209784304258e-05,
+      "loss": 3.8893,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.110371189553206e-05,
+      "loss": 3.8963,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.109534232682527e-05,
+      "loss": 3.9061,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.108695637931475e-05,
+      "loss": 3.8996,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.107857043180423e-05,
+      "loss": 3.8931,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1070184484293713e-05,
+      "loss": 3.8944,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1061798536783193e-05,
+      "loss": 3.8974,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.105341258927267e-05,
+      "loss": 3.889,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.104502664176215e-05,
+      "loss": 3.897,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.103664069425163e-05,
+      "loss": 3.8878,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1028271125544842e-05,
+      "loss": 3.885,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1019885178034322e-05,
+      "loss": 3.8899,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1011499230523802e-05,
+      "loss": 3.8936,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1003113283013282e-05,
+      "loss": 3.8902,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.099474371430649e-05,
+      "loss": 3.8954,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.098635776679597e-05,
+      "loss": 3.8987,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0977988198089184e-05,
+      "loss": 3.9039,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0969618629382396e-05,
+      "loss": 3.8866,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0961232681871876e-05,
+      "loss": 3.8964,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0952846734361356e-05,
+      "loss": 3.8972,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0944460786850836e-05,
+      "loss": 3.8859,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0936074839340312e-05,
+      "loss": 3.8806,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0927688891829796e-05,
+      "loss": 3.8852,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0919302944319276e-05,
+      "loss": 3.893,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0910916996808756e-05,
+      "loss": 3.8966,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0902547428101965e-05,
+      "loss": 3.8857,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0894161480591445e-05,
+      "loss": 3.8885,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0885775533080925e-05,
+      "loss": 3.9084,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0877389585570405e-05,
+      "loss": 3.8954,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0869003638059885e-05,
+      "loss": 3.8973,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0860634069353097e-05,
+      "loss": 3.8844,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0852248121842577e-05,
+      "loss": 3.8746,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0843862174332057e-05,
+      "loss": 3.9054,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0835476226821537e-05,
+      "loss": 3.8889,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0827090279311017e-05,
+      "loss": 3.8861,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0818704331800497e-05,
+      "loss": 3.8948,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0810318384289977e-05,
+      "loss": 3.8886,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0801932436779457e-05,
+      "loss": 3.8802,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.079356286807267e-05,
+      "loss": 3.8856,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0785176920562146e-05,
+      "loss": 3.8845,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0776790973051626e-05,
+      "loss": 3.8844,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0768405025541106e-05,
+      "loss": 3.8992,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.076001907803059e-05,
+      "loss": 3.8863,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.075163313052007e-05,
+      "loss": 3.8828,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.074324718300955e-05,
+      "loss": 3.8974,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.073486123549903e-05,
+      "loss": 3.8784,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0726491666792238e-05,
+      "loss": 3.8868,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.071812209808545e-05,
+      "loss": 3.8872,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.070973615057493e-05,
+      "loss": 3.8865,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.070135020306441e-05,
+      "loss": 3.8862,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.069296425555389e-05,
+      "loss": 3.8874,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.068457830804337e-05,
+      "loss": 3.8708,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.067619236053285e-05,
+      "loss": 3.8849,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0667806413022327e-05,
+      "loss": 3.9034,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0659436844315543e-05,
+      "loss": 3.8827,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0651050896805023e-05,
+      "loss": 3.9056,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.06426649492945e-05,
+      "loss": 3.8962,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.063427900178398e-05,
+      "loss": 3.8976,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.062589305427346e-05,
+      "loss": 3.8827,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.061750710676294e-05,
+      "loss": 3.8971,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060912115925242e-05,
+      "loss": 3.8935,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.06007352117419e-05,
+      "loss": 3.8809,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.059236564303511e-05,
+      "loss": 3.8982,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0583996074328324e-05,
+      "loss": 3.8905,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.05756101268178e-05,
+      "loss": 3.8985,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.056722417930728e-05,
+      "loss": 3.8956,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055883823179676e-05,
+      "loss": 3.8772,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0550452284286244e-05,
+      "loss": 3.8768,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0542066336775724e-05,
+      "loss": 3.8818,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0533696768068933e-05,
+      "loss": 3.8934,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0525310820558413e-05,
+      "loss": 3.8877,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0516924873047893e-05,
+      "loss": 3.8892,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0508538925537373e-05,
+      "loss": 3.8891,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0500152978026853e-05,
+      "loss": 3.8824,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0491767030516333e-05,
+      "loss": 3.8839,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0483381083005813e-05,
+      "loss": 3.8851,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0474995135495293e-05,
+      "loss": 3.899,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0466609187984773e-05,
+      "loss": 3.9043,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0458239619277982e-05,
+      "loss": 3.8899,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.044985367176746e-05,
+      "loss": 3.884,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0441467724256945e-05,
+      "loss": 3.8897,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0433098155550154e-05,
+      "loss": 3.8879,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0424712208039634e-05,
+      "loss": 3.9005,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0416326260529114e-05,
+      "loss": 3.888,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0407940313018594e-05,
+      "loss": 3.8867,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0399554365508074e-05,
+      "loss": 3.8922,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0391168417997554e-05,
+      "loss": 3.8931,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0382782470487037e-05,
+      "loss": 3.8795,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0374412901780246e-05,
+      "loss": 3.8894,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0366026954269726e-05,
+      "loss": 3.8845,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0357641006759206e-05,
+      "loss": 3.8776,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0349255059248686e-05,
+      "loss": 3.8794,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.03408854905419e-05,
+      "loss": 3.8888,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.033249954303138e-05,
+      "loss": 3.8906,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.032411359552086e-05,
+      "loss": 3.8822,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0315727648010335e-05,
+      "loss": 3.8862,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0307341700499815e-05,
+      "loss": 3.8848,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0298955752989295e-05,
+      "loss": 3.8968,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0290569805478775e-05,
+      "loss": 3.9054,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0282183857968255e-05,
+      "loss": 3.8965,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.027379791045774e-05,
+      "loss": 3.8868,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0265411962947218e-05,
+      "loss": 3.8838,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0257042394240427e-05,
+      "loss": 3.8899,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0248656446729907e-05,
+      "loss": 3.8766,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0240303256826852e-05,
+      "loss": 3.8886,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0231917309316332e-05,
+      "loss": 3.8875,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.022353136180581e-05,
+      "loss": 3.8841,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.021514541429529e-05,
+      "loss": 3.8891,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.020675946678477e-05,
+      "loss": 3.8875,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.019838989807798e-05,
+      "loss": 3.8948,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.019000395056746e-05,
+      "loss": 3.8878,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.018161800305694e-05,
+      "loss": 3.8852,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.017323205554642e-05,
+      "loss": 3.8921,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.01648461080359e-05,
+      "loss": 3.8943,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.015646016052538e-05,
+      "loss": 3.9001,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.014807421301486e-05,
+      "loss": 3.8849,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013968826550434e-05,
+      "loss": 3.8972,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013130231799382e-05,
+      "loss": 3.8913,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.01229163704833e-05,
+      "loss": 3.8963,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.011453042297278e-05,
+      "loss": 3.8882,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.010614447546226e-05,
+      "loss": 3.8912,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.009777490675547e-05,
+      "loss": 3.8849,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.008938895924495e-05,
+      "loss": 3.8971,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.008100301173443e-05,
+      "loss": 3.8786,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.007261706422391e-05,
+      "loss": 3.8876,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0064247495517122e-05,
+      "loss": 3.8897,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0055861548006602e-05,
+      "loss": 3.8931,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0047475600496082e-05,
+      "loss": 3.8887,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0039089652985562e-05,
+      "loss": 3.8902,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0030720084278775e-05,
+      "loss": 3.8977,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0022334136768255e-05,
+      "loss": 3.8807,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0013948189257734e-05,
+      "loss": 3.8972,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9863340854644775,
+      "eval_runtime": 297.0737,
+      "eval_samples_per_second": 1284.5,
+      "eval_steps_per_second": 40.142,
+      "step": 1831668
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0005562241747214e-05,
+      "loss": 3.8819,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9997176294236694e-05,
+      "loss": 3.8779,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.998879034672617e-05,
+      "loss": 3.8903,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.998040439921565e-05,
+      "loss": 3.8909,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.997201845170513e-05,
+      "loss": 3.8986,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9963632504194614e-05,
+      "loss": 3.8855,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9955246556684094e-05,
+      "loss": 3.8875,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9946860609173574e-05,
+      "loss": 3.8857,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9938474661663054e-05,
+      "loss": 3.8849,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.993008871415253e-05,
+      "loss": 3.883,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.992170276664201e-05,
+      "loss": 3.8947,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.991331681913149e-05,
+      "loss": 3.897,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9904947250424703e-05,
+      "loss": 3.8822,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9896561302914183e-05,
+      "loss": 3.879,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9888175355403663e-05,
+      "loss": 3.896,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9879789407893143e-05,
+      "loss": 3.8735,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9871403460382623e-05,
+      "loss": 3.8867,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9863017512872103e-05,
+      "loss": 3.8832,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9854631565361583e-05,
+      "loss": 3.8913,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9846245617851063e-05,
+      "loss": 3.8952,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9837859670340543e-05,
+      "loss": 3.8945,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9829490101633755e-05,
+      "loss": 3.8891,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9821104154123235e-05,
+      "loss": 3.8856,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.981271820661271e-05,
+      "loss": 3.8901,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.980433225910219e-05,
+      "loss": 3.8864,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.979594631159167e-05,
+      "loss": 3.8916,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9787560364081155e-05,
+      "loss": 3.8793,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9779190795374364e-05,
+      "loss": 3.8803,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9770804847863844e-05,
+      "loss": 3.8879,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9762418900353324e-05,
+      "loss": 3.8835,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9754032952842804e-05,
+      "loss": 3.8825,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9745647005332284e-05,
+      "loss": 3.8909,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9737261057821764e-05,
+      "loss": 3.8954,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9728875110311244e-05,
+      "loss": 3.8928,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9720505541604456e-05,
+      "loss": 3.8818,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9712119594093936e-05,
+      "loss": 3.8928,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9703733646583416e-05,
+      "loss": 3.8905,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9695347699072893e-05,
+      "loss": 3.8825,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9686961751562376e-05,
+      "loss": 3.8778,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9678575804051856e-05,
+      "loss": 3.8774,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9670189856541336e-05,
+      "loss": 3.8836,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9661803909030816e-05,
+      "loss": 3.8935,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9653417961520296e-05,
+      "loss": 3.8779,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9645048392813505e-05,
+      "loss": 3.8806,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9636662445302985e-05,
+      "loss": 3.9021,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9628276497792465e-05,
+      "loss": 3.8858,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9619906929085677e-05,
+      "loss": 3.8942,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9611520981575157e-05,
+      "loss": 3.8806,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9603135034064637e-05,
+      "loss": 3.8626,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9594765465357846e-05,
+      "loss": 3.9,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9586379517847326e-05,
+      "loss": 3.8862,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.957799357033681e-05,
+      "loss": 3.8822,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.956962400163002e-05,
+      "loss": 3.8841,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95612380541195e-05,
+      "loss": 3.883,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.955285210660898e-05,
+      "loss": 3.8742,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.954446615909846e-05,
+      "loss": 3.877,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.953608021158794e-05,
+      "loss": 3.8805,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.952769426407742e-05,
+      "loss": 3.8757,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95193083165669e-05,
+      "loss": 3.8915,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951092236905638e-05,
+      "loss": 3.8795,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.950253642154586e-05,
+      "loss": 3.877,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.949415047403534e-05,
+      "loss": 3.8931,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.948576452652482e-05,
+      "loss": 3.8664,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.94773785790143e-05,
+      "loss": 3.8793,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.946900901030751e-05,
+      "loss": 3.8787,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.946062306279699e-05,
+      "loss": 3.8855,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.945223711528647e-05,
+      "loss": 3.8783,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.944385116777595e-05,
+      "loss": 3.883,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9435465220265427e-05,
+      "loss": 3.8595,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.942709565155864e-05,
+      "loss": 3.8791,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9418709704048123e-05,
+      "loss": 3.8927,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9410323756537603e-05,
+      "loss": 3.8767,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.940193780902708e-05,
+      "loss": 3.8986,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.939355186151656e-05,
+      "loss": 3.8892,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.938516591400604e-05,
+      "loss": 3.8975,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937677996649552e-05,
+      "loss": 3.8742,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9368394018985e-05,
+      "loss": 3.8895,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9360024450278212e-05,
+      "loss": 3.8851,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9351638502767692e-05,
+      "loss": 3.8749,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.93432689340609e-05,
+      "loss": 3.8932,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.933488298655038e-05,
+      "loss": 3.8821,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.932649703903986e-05,
+      "loss": 3.8927,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.931811109152934e-05,
+      "loss": 3.8883,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9309725144018824e-05,
+      "loss": 3.8701,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9301339196508304e-05,
+      "loss": 3.8725,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9292953248997784e-05,
+      "loss": 3.8744,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.928456730148726e-05,
+      "loss": 3.8874,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9276197732780473e-05,
+      "loss": 3.8797,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9267828164073686e-05,
+      "loss": 3.879,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9259442216563166e-05,
+      "loss": 3.881,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9251056269052646e-05,
+      "loss": 3.8791,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9242670321542125e-05,
+      "loss": 3.8781,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9234284374031605e-05,
+      "loss": 3.8745,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9225914805324815e-05,
+      "loss": 3.8979,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9217528857814294e-05,
+      "loss": 3.8945,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9209142910303778e-05,
+      "loss": 3.8869,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9200756962793258e-05,
+      "loss": 3.8742,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9192387394086467e-05,
+      "loss": 3.884,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9184001446575947e-05,
+      "loss": 3.8782,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9175615499065427e-05,
+      "loss": 3.8966,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9167229551554903e-05,
+      "loss": 3.8817,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9158843604044387e-05,
+      "loss": 3.8816,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9150457656533867e-05,
+      "loss": 3.8863,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9142071709023347e-05,
+      "loss": 3.8871,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9133702140316556e-05,
+      "loss": 3.8723,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9125316192806036e-05,
+      "loss": 3.8827,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9116930245295516e-05,
+      "loss": 3.8758,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9108544297784996e-05,
+      "loss": 3.8738,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.910015835027448e-05,
+      "loss": 3.8743,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.909177240276396e-05,
+      "loss": 3.8792,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.908338645525344e-05,
+      "loss": 3.882,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9075016886546648e-05,
+      "loss": 3.8813,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9066630939036128e-05,
+      "loss": 3.8794,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9058244991525608e-05,
+      "loss": 3.8751,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9049859044015088e-05,
+      "loss": 3.8872,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.90414894753083e-05,
+      "loss": 3.899,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.903310352779778e-05,
+      "loss": 3.887,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.902471758028726e-05,
+      "loss": 3.8846,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9016331632776737e-05,
+      "loss": 3.873,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9007945685266217e-05,
+      "loss": 3.8839,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8999559737755697e-05,
+      "loss": 3.877,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.899120654785264e-05,
+      "loss": 3.8807,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.898282060034212e-05,
+      "loss": 3.8748,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.897445103163533e-05,
+      "loss": 3.88,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896606508412481e-05,
+      "loss": 3.8832,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8957679136614294e-05,
+      "loss": 3.8809,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8949293189103774e-05,
+      "loss": 3.8864,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8940907241593254e-05,
+      "loss": 3.8815,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8932521294082734e-05,
+      "loss": 3.8742,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.892413534657221e-05,
+      "loss": 3.8893,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.891574939906169e-05,
+      "loss": 3.8876,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890736345155117e-05,
+      "loss": 3.8956,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889897750404065e-05,
+      "loss": 3.8762,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8890591556530134e-05,
+      "loss": 3.893,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8882205609019614e-05,
+      "loss": 3.8841,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.887381966150909e-05,
+      "loss": 3.8875,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.886543371399857e-05,
+      "loss": 3.8855,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.885704776648805e-05,
+      "loss": 3.8834,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.884866181897753e-05,
+      "loss": 3.877,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8840292250270743e-05,
+      "loss": 3.8916,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8831906302760223e-05,
+      "loss": 3.8716,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8823520355249702e-05,
+      "loss": 3.8811,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8815134407739182e-05,
+      "loss": 3.8861,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8806748460228662e-05,
+      "loss": 3.8891,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8798362512718142e-05,
+      "loss": 3.8791,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8789976565207622e-05,
+      "loss": 3.8854,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8781606996500835e-05,
+      "loss": 3.8886,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8773221048990315e-05,
+      "loss": 3.879,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8764835101479795e-05,
+      "loss": 3.8877,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.983210802078247,
+      "eval_runtime": 295.0879,
+      "eval_samples_per_second": 1293.143,
+      "eval_steps_per_second": 40.412,
+      "step": 1907988
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.875644915396927e-05,
+      "loss": 3.8742,
+      "step": 1908224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.874806320645875e-05,
+      "loss": 3.8762,
+      "step": 1908736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8739693637751964e-05,
+      "loss": 3.8815,
+      "step": 1909248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8731307690241444e-05,
+      "loss": 3.8861,
+      "step": 1909760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8722921742730924e-05,
+      "loss": 3.8925,
+      "step": 1910272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8714535795220404e-05,
+      "loss": 3.8782,
+      "step": 1910784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8706149847709884e-05,
+      "loss": 3.8836,
+      "step": 1911296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8697780279003096e-05,
+      "loss": 3.8773,
+      "step": 1911808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8689394331492573e-05,
+      "loss": 3.8754,
+      "step": 1912320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8681008383982056e-05,
+      "loss": 3.8795,
+      "step": 1912832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8672622436471536e-05,
+      "loss": 3.8876,
+      "step": 1913344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8664252867764745e-05,
+      "loss": 3.8888,
+      "step": 1913856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8655883299057957e-05,
+      "loss": 3.8837,
+      "step": 1914368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8647497351547437e-05,
+      "loss": 3.8666,
+      "step": 1914880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8639111404036917e-05,
+      "loss": 3.8867,
+      "step": 1915392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8630725456526397e-05,
+      "loss": 3.8692,
+      "step": 1915904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8622339509015877e-05,
+      "loss": 3.8812,
+      "step": 1916416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8613953561505357e-05,
+      "loss": 3.8786,
+      "step": 1916928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8605567613994837e-05,
+      "loss": 3.8799,
+      "step": 1917440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8597181666484317e-05,
+      "loss": 3.8886,
+      "step": 1917952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8588795718973797e-05,
+      "loss": 3.8925,
+      "step": 1918464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8580409771463277e-05,
+      "loss": 3.8854,
+      "step": 1918976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8572023823952757e-05,
+      "loss": 3.8807,
+      "step": 1919488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.856365425524597e-05,
+      "loss": 3.8785,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.855526830773545e-05,
+      "loss": 3.8848,
+      "step": 1920512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8546882360224926e-05,
+      "loss": 3.8808,
+      "step": 1921024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853851279151814e-05,
+      "loss": 3.8804,
+      "step": 1921536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853012684400762e-05,
+      "loss": 3.875,
+      "step": 1922048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8521740896497102e-05,
+      "loss": 3.8767,
+      "step": 1922560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.851335494898658e-05,
+      "loss": 3.8813,
+      "step": 1923072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.850496900147606e-05,
+      "loss": 3.87,
+      "step": 1923584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8496583053965538e-05,
+      "loss": 3.8884,
+      "step": 1924096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8488197106455018e-05,
+      "loss": 3.8866,
+      "step": 1924608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8479811158944498e-05,
+      "loss": 3.886,
+      "step": 1925120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.847144159023771e-05,
+      "loss": 3.8761,
+      "step": 1925632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8463072021530923e-05,
+      "loss": 3.8852,
+      "step": 1926144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.84546860740204e-05,
+      "loss": 3.8844,
+      "step": 1926656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.844630012650988e-05,
+      "loss": 3.8748,
+      "step": 1927168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.843791417899936e-05,
+      "loss": 3.8729,
+      "step": 1927680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8429544610292572e-05,
+      "loss": 3.8733,
+      "step": 1928192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8421175041585785e-05,
+      "loss": 3.876,
+      "step": 1928704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8412789094075265e-05,
+      "loss": 3.8848,
+      "step": 1929216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8404403146564745e-05,
+      "loss": 3.8767,
+      "step": 1929728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8396017199054224e-05,
+      "loss": 3.8744,
+      "step": 1930240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.83876312515437e-05,
+      "loss": 3.8937,
+      "step": 1930752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.837924530403318e-05,
+      "loss": 3.8806,
+      "step": 1931264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8370859356522664e-05,
+      "loss": 3.8884,
+      "step": 1931776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8362473409012144e-05,
+      "loss": 3.8765,
+      "step": 1932288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8354087461501624e-05,
+      "loss": 3.8579,
+      "step": 1932800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8345701513991104e-05,
+      "loss": 3.8941,
+      "step": 1933312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.833731556648058e-05,
+      "loss": 3.8763,
+      "step": 1933824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8328945997773793e-05,
+      "loss": 3.8826,
+      "step": 1934336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8320560050263273e-05,
+      "loss": 3.8757,
+      "step": 1934848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8312174102752757e-05,
+      "loss": 3.8792,
+      "step": 1935360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8303788155242233e-05,
+      "loss": 3.8694,
+      "step": 1935872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8295402207731713e-05,
+      "loss": 3.8683,
+      "step": 1936384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8287016260221193e-05,
+      "loss": 3.8743,
+      "step": 1936896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8278630312710673e-05,
+      "loss": 3.8717,
+      "step": 1937408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8270244365200153e-05,
+      "loss": 3.8875,
+      "step": 1937920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8261891175297098e-05,
+      "loss": 3.8742,
+      "step": 1938432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8253505227786578e-05,
+      "loss": 3.8714,
+      "step": 1938944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8245119280276054e-05,
+      "loss": 3.8908,
+      "step": 1939456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8236733332765534e-05,
+      "loss": 3.8617,
+      "step": 1939968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8228347385255014e-05,
+      "loss": 3.8684,
+      "step": 1940480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8219961437744494e-05,
+      "loss": 3.8807,
+      "step": 1940992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8211575490233974e-05,
+      "loss": 3.878,
+      "step": 1941504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8203189542723458e-05,
+      "loss": 3.8732,
+      "step": 1942016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8194803595212938e-05,
+      "loss": 3.8783,
+      "step": 1942528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8186417647702414e-05,
+      "loss": 3.8568,
+      "step": 1943040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8178048078995627e-05,
+      "loss": 3.8688,
+      "step": 1943552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8169662131485107e-05,
+      "loss": 3.8906,
+      "step": 1944064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8161276183974583e-05,
+      "loss": 3.8712,
+      "step": 1944576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8152890236464067e-05,
+      "loss": 3.8899,
+      "step": 1945088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8144504288953546e-05,
+      "loss": 3.8841,
+      "step": 1945600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8136118341443026e-05,
+      "loss": 3.8919,
+      "step": 1946112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8127748772736236e-05,
+      "loss": 3.8682,
+      "step": 1946624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8119362825225715e-05,
+      "loss": 3.8852,
+      "step": 1947136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8110976877715195e-05,
+      "loss": 3.882,
+      "step": 1947648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.810260730900841e-05,
+      "loss": 3.8658,
+      "step": 1948160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8094221361497888e-05,
+      "loss": 3.8903,
+      "step": 1948672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8085835413987368e-05,
+      "loss": 3.8782,
+      "step": 1949184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8077449466476848e-05,
+      "loss": 3.8803,
+      "step": 1949696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8069063518966328e-05,
+      "loss": 3.8886,
+      "step": 1950208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8060677571455808e-05,
+      "loss": 3.8661,
+      "step": 1950720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8052291623945288e-05,
+      "loss": 3.8656,
+      "step": 1951232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8043905676434768e-05,
+      "loss": 3.8702,
+      "step": 1951744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8035519728924248e-05,
+      "loss": 3.8804,
+      "step": 1952256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8027133781413728e-05,
+      "loss": 3.8771,
+      "step": 1952768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.801876421270694e-05,
+      "loss": 3.8743,
+      "step": 1953280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8010378265196417e-05,
+      "loss": 3.8757,
+      "step": 1953792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8001992317685897e-05,
+      "loss": 3.8777,
+      "step": 1954304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.799360637017538e-05,
+      "loss": 3.8694,
+      "step": 1954816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.798522042266486e-05,
+      "loss": 3.8718,
+      "step": 1955328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.797683447515434e-05,
+      "loss": 3.8923,
+      "step": 1955840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796844852764382e-05,
+      "loss": 3.8884,
+      "step": 1956352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.79600625801333e-05,
+      "loss": 3.8829,
+      "step": 1956864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.795169301142651e-05,
+      "loss": 3.8682,
+      "step": 1957376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.794332344271972e-05,
+      "loss": 3.8789,
+      "step": 1957888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.79349374952092e-05,
+      "loss": 3.8709,
+      "step": 1958400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.792655154769868e-05,
+      "loss": 3.8906,
+      "step": 1958912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.791816560018816e-05,
+      "loss": 3.8794,
+      "step": 1959424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.790979603148137e-05,
+      "loss": 3.8727,
+      "step": 1959936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.790141008397085e-05,
+      "loss": 3.8792,
+      "step": 1960448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.789302413646033e-05,
+      "loss": 3.8856,
+      "step": 1960960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7884654567753543e-05,
+      "loss": 3.8672,
+      "step": 1961472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7876268620243023e-05,
+      "loss": 3.8785,
+      "step": 1961984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7867882672732503e-05,
+      "loss": 3.8661,
+      "step": 1962496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7859496725221982e-05,
+      "loss": 3.8714,
+      "step": 1963008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7851110777711462e-05,
+      "loss": 3.8693,
+      "step": 1963520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7842724830200942e-05,
+      "loss": 3.8716,
+      "step": 1964032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7834338882690422e-05,
+      "loss": 3.8757,
+      "step": 1964544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7825952935179902e-05,
+      "loss": 3.8812,
+      "step": 1965056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7817566987669382e-05,
+      "loss": 3.8719,
+      "step": 1965568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7809181040158862e-05,
+      "loss": 3.871,
+      "step": 1966080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7800795092648342e-05,
+      "loss": 3.8838,
+      "step": 1966592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7792409145137822e-05,
+      "loss": 3.8922,
+      "step": 1967104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7784023197627302e-05,
+      "loss": 3.8802,
+      "step": 1967616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7775637250116782e-05,
+      "loss": 3.883,
+      "step": 1968128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7767267681409995e-05,
+      "loss": 3.8689,
+      "step": 1968640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7758881733899474e-05,
+      "loss": 3.8784,
+      "step": 1969152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7750495786388954e-05,
+      "loss": 3.8749,
+      "step": 1969664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7742126217682164e-05,
+      "loss": 3.8713,
+      "step": 1970176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7733740270171644e-05,
+      "loss": 3.8744,
+      "step": 1970688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7725354322661127e-05,
+      "loss": 3.877,
+      "step": 1971200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7716968375150603e-05,
+      "loss": 3.8769,
+      "step": 1971712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7708582427640083e-05,
+      "loss": 3.8737,
+      "step": 1972224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7700212858933296e-05,
+      "loss": 3.8809,
+      "step": 1972736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7691826911422776e-05,
+      "loss": 3.8782,
+      "step": 1973248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7683440963912252e-05,
+      "loss": 3.8709,
+      "step": 1973760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7675055016401736e-05,
+      "loss": 3.8799,
+      "step": 1974272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7666669068891216e-05,
+      "loss": 3.8872,
+      "step": 1974784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7658283121380696e-05,
+      "loss": 3.8865,
+      "step": 1975296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7649897173870176e-05,
+      "loss": 3.8747,
+      "step": 1975808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7641511226359656e-05,
+      "loss": 3.8887,
+      "step": 1976320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7633141657652865e-05,
+      "loss": 3.8839,
+      "step": 1976832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7624755710142345e-05,
+      "loss": 3.8813,
+      "step": 1977344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7616386141435557e-05,
+      "loss": 3.8821,
+      "step": 1977856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7608000193925037e-05,
+      "loss": 3.8768,
+      "step": 1978368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7599614246414517e-05,
+      "loss": 3.8741,
+      "step": 1978880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7591228298903997e-05,
+      "loss": 3.884,
+      "step": 1979392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7582842351393477e-05,
+      "loss": 3.8683,
+      "step": 1979904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7574456403882957e-05,
+      "loss": 3.8765,
+      "step": 1980416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7566070456372437e-05,
+      "loss": 3.884,
+      "step": 1980928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.755770088766565e-05,
+      "loss": 3.8817,
+      "step": 1981440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.754931494015513e-05,
+      "loss": 3.8734,
+      "step": 1981952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7540928992644606e-05,
+      "loss": 3.8783,
+      "step": 1982464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7532543045134086e-05,
+      "loss": 3.8841,
+      "step": 1982976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7524157097623566e-05,
+      "loss": 3.8783,
+      "step": 1983488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7515771150113046e-05,
+      "loss": 3.8826,
+      "step": 1984000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.981627941131592,
+      "eval_runtime": 295.6949,
+      "eval_samples_per_second": 1290.489,
+      "eval_steps_per_second": 40.329,
+      "step": 1984308
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.750738520260253e-05,
+      "loss": 3.8592,
+      "step": 1984512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.749899925509201e-05,
+      "loss": 3.8715,
+      "step": 1985024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.749061330758149e-05,
+      "loss": 3.8793,
+      "step": 1985536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7482227360070966e-05,
+      "loss": 3.8791,
+      "step": 1986048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7473841412560445e-05,
+      "loss": 3.8894,
+      "step": 1986560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7465455465049925e-05,
+      "loss": 3.8738,
+      "step": 1987072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7457069517539405e-05,
+      "loss": 3.8798,
+      "step": 1987584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7448683570028885e-05,
+      "loss": 3.873,
+      "step": 1988096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.744029762251837e-05,
+      "loss": 3.8721,
+      "step": 1988608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.743191167500785e-05,
+      "loss": 3.8738,
+      "step": 1989120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7423525727497325e-05,
+      "loss": 3.8813,
+      "step": 1989632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7415139779986805e-05,
+      "loss": 3.8849,
+      "step": 1990144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7406770211280018e-05,
+      "loss": 3.8827,
+      "step": 1990656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7398384263769498e-05,
+      "loss": 3.8582,
+      "step": 1991168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7389998316258978e-05,
+      "loss": 3.8833,
+      "step": 1991680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7381612368748458e-05,
+      "loss": 3.862,
+      "step": 1992192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7373226421237937e-05,
+      "loss": 3.8763,
+      "step": 1992704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7364840473727417e-05,
+      "loss": 3.8721,
+      "step": 1993216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7356454526216897e-05,
+      "loss": 3.8745,
+      "step": 1993728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7348068578706377e-05,
+      "loss": 3.8821,
+      "step": 1994240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7339682631195854e-05,
+      "loss": 3.8888,
+      "step": 1994752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7331296683685337e-05,
+      "loss": 3.881,
+      "step": 1995264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7322910736174817e-05,
+      "loss": 3.8757,
+      "step": 1995776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.731454116746803e-05,
+      "loss": 3.8716,
+      "step": 1996288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7306155219957506e-05,
+      "loss": 3.8817,
+      "step": 1996800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7297769272446986e-05,
+      "loss": 3.8793,
+      "step": 1997312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7289383324936466e-05,
+      "loss": 3.8721,
+      "step": 1997824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.728101375622968e-05,
+      "loss": 3.8728,
+      "step": 1998336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.727264418752289e-05,
+      "loss": 3.8731,
+      "step": 1998848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.726425824001237e-05,
+      "loss": 3.8757,
+      "step": 1999360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.725587229250185e-05,
+      "loss": 3.8651,
+      "step": 1999872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7247486344991328e-05,
+      "loss": 3.8859,
+      "step": 2000384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7239100397480808e-05,
+      "loss": 3.8836,
+      "step": 2000896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.723071444997029e-05,
+      "loss": 3.8791,
+      "step": 2001408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.722232850245977e-05,
+      "loss": 3.8742,
+      "step": 2001920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.721395893375298e-05,
+      "loss": 3.8752,
+      "step": 2002432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.720557298624246e-05,
+      "loss": 3.8827,
+      "step": 2002944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.719718703873194e-05,
+      "loss": 3.8718,
+      "step": 2003456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.718880109122142e-05,
+      "loss": 3.8678,
+      "step": 2003968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.71804151437109e-05,
+      "loss": 3.8683,
+      "step": 2004480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7172029196200383e-05,
+      "loss": 3.8712,
+      "step": 2004992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.716364324868986e-05,
+      "loss": 3.8773,
+      "step": 2005504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.715525730117934e-05,
+      "loss": 3.8682,
+      "step": 2006016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.714687135366882e-05,
+      "loss": 3.8733,
+      "step": 2006528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.71384854061583e-05,
+      "loss": 3.8868,
+      "step": 2007040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.713011583745151e-05,
+      "loss": 3.8738,
+      "step": 2007552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7121729889940992e-05,
+      "loss": 3.8805,
+      "step": 2008064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7113343942430472e-05,
+      "loss": 3.8758,
+      "step": 2008576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7104957994919952e-05,
+      "loss": 3.8535,
+      "step": 2009088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7096572047409432e-05,
+      "loss": 3.8877,
+      "step": 2009600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.708820247870264e-05,
+      "loss": 3.8722,
+      "step": 2010112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.707981653119212e-05,
+      "loss": 3.8776,
+      "step": 2010624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.70714305836816e-05,
+      "loss": 3.8759,
+      "step": 2011136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7063044636171084e-05,
+      "loss": 3.8717,
+      "step": 2011648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7054658688660564e-05,
+      "loss": 3.8628,
+      "step": 2012160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7046289119953773e-05,
+      "loss": 3.8677,
+      "step": 2012672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7037903172443253e-05,
+      "loss": 3.8696,
+      "step": 2013184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7029517224932733e-05,
+      "loss": 3.8673,
+      "step": 2013696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7021131277422213e-05,
+      "loss": 3.8826,
+      "step": 2014208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7012761708715426e-05,
+      "loss": 3.8699,
+      "step": 2014720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7004375761204906e-05,
+      "loss": 3.8655,
+      "step": 2015232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6995989813694386e-05,
+      "loss": 3.8866,
+      "step": 2015744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6987603866183862e-05,
+      "loss": 3.858,
+      "step": 2016256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6979217918673342e-05,
+      "loss": 3.8625,
+      "step": 2016768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6970831971162822e-05,
+      "loss": 3.8787,
+      "step": 2017280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6962446023652302e-05,
+      "loss": 3.8709,
+      "step": 2017792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6954060076141785e-05,
+      "loss": 3.8662,
+      "step": 2018304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6945674128631265e-05,
+      "loss": 3.8753,
+      "step": 2018816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6937288181120745e-05,
+      "loss": 3.8565,
+      "step": 2019328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6928902233610222e-05,
+      "loss": 3.8631,
+      "step": 2019840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6920516286099702e-05,
+      "loss": 3.8833,
+      "step": 2020352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6912146717392914e-05,
+      "loss": 3.8673,
+      "step": 2020864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6903760769882394e-05,
+      "loss": 3.8834,
+      "step": 2021376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6895374822371874e-05,
+      "loss": 3.8762,
+      "step": 2021888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6886988874861354e-05,
+      "loss": 3.891,
+      "step": 2022400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6878619306154567e-05,
+      "loss": 3.8629,
+      "step": 2022912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6870233358644043e-05,
+      "loss": 3.8835,
+      "step": 2023424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6861863789937256e-05,
+      "loss": 3.8793,
+      "step": 2023936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.685347784242674e-05,
+      "loss": 3.8596,
+      "step": 2024448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.684509189491622e-05,
+      "loss": 3.8863,
+      "step": 2024960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6836705947405695e-05,
+      "loss": 3.8728,
+      "step": 2025472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6828319999895175e-05,
+      "loss": 3.8764,
+      "step": 2025984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6819934052384655e-05,
+      "loss": 3.8828,
+      "step": 2026496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6811548104874135e-05,
+      "loss": 3.8629,
+      "step": 2027008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6803178536167348e-05,
+      "loss": 3.8598,
+      "step": 2027520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6794792588656828e-05,
+      "loss": 3.8632,
+      "step": 2028032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6786406641146308e-05,
+      "loss": 3.8761,
+      "step": 2028544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6778020693635788e-05,
+      "loss": 3.8705,
+      "step": 2029056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6769651124928997e-05,
+      "loss": 3.868,
+      "step": 2029568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6761265177418477e-05,
+      "loss": 3.8682,
+      "step": 2030080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.675287922990796e-05,
+      "loss": 3.8743,
+      "step": 2030592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.674449328239744e-05,
+      "loss": 3.8703,
+      "step": 2031104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.673610733488692e-05,
+      "loss": 3.8659,
+      "step": 2031616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.67277213873764e-05,
+      "loss": 3.8854,
+      "step": 2032128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6719335439865877e-05,
+      "loss": 3.8819,
+      "step": 2032640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6710949492355356e-05,
+      "loss": 3.8791,
+      "step": 2033152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.670257992364857e-05,
+      "loss": 3.8637,
+      "step": 2033664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.669421035494178e-05,
+      "loss": 3.8714,
+      "step": 2034176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.668582440743126e-05,
+      "loss": 3.8685,
+      "step": 2034688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.667743845992074e-05,
+      "loss": 3.8853,
+      "step": 2035200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.666905251241022e-05,
+      "loss": 3.8729,
+      "step": 2035712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6660666564899698e-05,
+      "loss": 3.8708,
+      "step": 2036224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.665229699619291e-05,
+      "loss": 3.8729,
+      "step": 2036736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6643911048682394e-05,
+      "loss": 3.8779,
+      "step": 2037248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6635525101171874e-05,
+      "loss": 3.8696,
+      "step": 2037760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.662713915366135e-05,
+      "loss": 3.8736,
+      "step": 2038272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.661875320615083e-05,
+      "loss": 3.8606,
+      "step": 2038784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.661036725864031e-05,
+      "loss": 3.8671,
+      "step": 2039296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.660198131112979e-05,
+      "loss": 3.8658,
+      "step": 2039808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.659359536361927e-05,
+      "loss": 3.864,
+      "step": 2040320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6585225794912483e-05,
+      "loss": 3.8695,
+      "step": 2040832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6576839847401963e-05,
+      "loss": 3.8725,
+      "step": 2041344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.656847027869517e-05,
+      "loss": 3.8699,
+      "step": 2041856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.656008433118465e-05,
+      "loss": 3.8659,
+      "step": 2042368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.655169838367413e-05,
+      "loss": 3.8802,
+      "step": 2042880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6543312436163615e-05,
+      "loss": 3.8849,
+      "step": 2043392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6534926488653095e-05,
+      "loss": 3.8769,
+      "step": 2043904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6526540541142575e-05,
+      "loss": 3.88,
+      "step": 2044416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6518154593632055e-05,
+      "loss": 3.8652,
+      "step": 2044928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.650976864612153e-05,
+      "loss": 3.871,
+      "step": 2045440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6501399077414744e-05,
+      "loss": 3.8719,
+      "step": 2045952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6493029508707956e-05,
+      "loss": 3.8654,
+      "step": 2046464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6484643561197436e-05,
+      "loss": 3.8705,
+      "step": 2046976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6476257613686916e-05,
+      "loss": 3.8725,
+      "step": 2047488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6467871666176396e-05,
+      "loss": 3.8766,
+      "step": 2048000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6459485718665876e-05,
+      "loss": 3.8701,
+      "step": 2048512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6451116149959085e-05,
+      "loss": 3.8754,
+      "step": 2049024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.644273020244857e-05,
+      "loss": 3.871,
+      "step": 2049536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.643434425493805e-05,
+      "loss": 3.8648,
+      "step": 2050048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.642595830742753e-05,
+      "loss": 3.8755,
+      "step": 2050560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6417572359917005e-05,
+      "loss": 3.8813,
+      "step": 2051072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6409186412406485e-05,
+      "loss": 3.8801,
+      "step": 2051584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6400816843699697e-05,
+      "loss": 3.8674,
+      "step": 2052096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6392430896189177e-05,
+      "loss": 3.8869,
+      "step": 2052608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6384044948678657e-05,
+      "loss": 3.8769,
+      "step": 2053120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6375659001168137e-05,
+      "loss": 3.8782,
+      "step": 2053632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6367273053657617e-05,
+      "loss": 3.8801,
+      "step": 2054144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6358887106147097e-05,
+      "loss": 3.8685,
+      "step": 2054656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6350517537440306e-05,
+      "loss": 3.8675,
+      "step": 2055168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6342131589929786e-05,
+      "loss": 3.8835,
+      "step": 2055680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.633374564241927e-05,
+      "loss": 3.8605,
+      "step": 2056192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.632535969490875e-05,
+      "loss": 3.8743,
+      "step": 2056704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.631697374739823e-05,
+      "loss": 3.8785,
+      "step": 2057216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.630858779988771e-05,
+      "loss": 3.8768,
+      "step": 2057728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.630021823118092e-05,
+      "loss": 3.8705,
+      "step": 2058240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.62918322836704e-05,
+      "loss": 3.8744,
+      "step": 2058752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.628344633615988e-05,
+      "loss": 3.8771,
+      "step": 2059264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.627506038864936e-05,
+      "loss": 3.8786,
+      "step": 2059776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.626667444113884e-05,
+      "loss": 3.8796,
+      "step": 2060288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.980487108230591,
+      "eval_runtime": 294.3922,
+      "eval_samples_per_second": 1296.2,
+      "eval_steps_per_second": 40.507,
+      "step": 2060628
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.625828849362832e-05,
+      "loss": 3.8515,
+      "step": 2060800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.62499025461178e-05,
+      "loss": 3.8701,
+      "step": 2061312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6241516598607278e-05,
+      "loss": 3.8705,
+      "step": 2061824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6233130651096758e-05,
+      "loss": 3.8738,
+      "step": 2062336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6224744703586238e-05,
+      "loss": 3.886,
+      "step": 2062848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6216358756075718e-05,
+      "loss": 3.8698,
+      "step": 2063360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6207972808565198e-05,
+      "loss": 3.8728,
+      "step": 2063872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6199586861054678e-05,
+      "loss": 3.8711,
+      "step": 2064384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6191200913544158e-05,
+      "loss": 3.8651,
+      "step": 2064896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6182814966033638e-05,
+      "loss": 3.8663,
+      "step": 2065408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6174429018523118e-05,
+      "loss": 3.8776,
+      "step": 2065920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6166043071012598e-05,
+      "loss": 3.8855,
+      "step": 2066432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.615767350230581e-05,
+      "loss": 3.8795,
+      "step": 2066944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.614928755479529e-05,
+      "loss": 3.8573,
+      "step": 2067456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.614090160728477e-05,
+      "loss": 3.8764,
+      "step": 2067968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6132515659774247e-05,
+      "loss": 3.8586,
+      "step": 2068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6124129712263727e-05,
+      "loss": 3.872,
+      "step": 2068992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6115743764753207e-05,
+      "loss": 3.8691,
+      "step": 2069504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6107357817242687e-05,
+      "loss": 3.8712,
+      "step": 2070016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.609897186973217e-05,
+      "loss": 3.8747,
+      "step": 2070528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.609060230102538e-05,
+      "loss": 3.8831,
+      "step": 2071040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.608221635351486e-05,
+      "loss": 3.8747,
+      "step": 2071552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.607383040600434e-05,
+      "loss": 3.8735,
+      "step": 2072064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.606544445849382e-05,
+      "loss": 3.8667,
+      "step": 2072576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.605707488978703e-05,
+      "loss": 3.877,
+      "step": 2073088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.604868894227651e-05,
+      "loss": 3.8742,
+      "step": 2073600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.604030299476599e-05,
+      "loss": 3.8672,
+      "step": 2074112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.603191704725547e-05,
+      "loss": 3.8689,
+      "step": 2074624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.602353109974495e-05,
+      "loss": 3.867,
+      "step": 2075136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6015145152234428e-05,
+      "loss": 3.8691,
+      "step": 2075648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6006759204723908e-05,
+      "loss": 3.8646,
+      "step": 2076160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5998373257213388e-05,
+      "loss": 3.8717,
+      "step": 2076672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.598998730970287e-05,
+      "loss": 3.8807,
+      "step": 2077184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.598160136219235e-05,
+      "loss": 3.8767,
+      "step": 2077696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.597321541468183e-05,
+      "loss": 3.8699,
+      "step": 2078208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.596484584597504e-05,
+      "loss": 3.8722,
+      "step": 2078720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.595645989846452e-05,
+      "loss": 3.8804,
+      "step": 2079232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5948073950954e-05,
+      "loss": 3.8679,
+      "step": 2079744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.593968800344348e-05,
+      "loss": 3.8628,
+      "step": 2080256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.593130205593296e-05,
+      "loss": 3.8682,
+      "step": 2080768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5922932487226172e-05,
+      "loss": 3.8639,
+      "step": 2081280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5914546539715652e-05,
+      "loss": 3.8725,
+      "step": 2081792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5906160592205132e-05,
+      "loss": 3.8643,
+      "step": 2082304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.589777464469461e-05,
+      "loss": 3.8684,
+      "step": 2082816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.588938869718409e-05,
+      "loss": 3.8775,
+      "step": 2083328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5881002749673572e-05,
+      "loss": 3.8773,
+      "step": 2083840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5872633180966785e-05,
+      "loss": 3.8772,
+      "step": 2084352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.586424723345626e-05,
+      "loss": 3.8682,
+      "step": 2084864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.585586128594574e-05,
+      "loss": 3.8485,
+      "step": 2085376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.584747533843522e-05,
+      "loss": 3.8789,
+      "step": 2085888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.58390893909247e-05,
+      "loss": 3.8661,
+      "step": 2086400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.583070344341418e-05,
+      "loss": 3.8742,
+      "step": 2086912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5822317495903664e-05,
+      "loss": 3.8724,
+      "step": 2087424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.581393154839314e-05,
+      "loss": 3.8685,
+      "step": 2087936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5805561979686354e-05,
+      "loss": 3.8577,
+      "step": 2088448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5797176032175833e-05,
+      "loss": 3.8581,
+      "step": 2088960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5788790084665313e-05,
+      "loss": 3.8679,
+      "step": 2089472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5780404137154793e-05,
+      "loss": 3.862,
+      "step": 2089984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5772018189644273e-05,
+      "loss": 3.8819,
+      "step": 2090496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5763632242133753e-05,
+      "loss": 3.8662,
+      "step": 2091008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5755246294623233e-05,
+      "loss": 3.8571,
+      "step": 2091520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5746860347112713e-05,
+      "loss": 3.8811,
+      "step": 2092032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5738490778405922e-05,
+      "loss": 3.858,
+      "step": 2092544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5730104830895402e-05,
+      "loss": 3.856,
+      "step": 2093056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5721718883384882e-05,
+      "loss": 3.875,
+      "step": 2093568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5713332935874366e-05,
+      "loss": 3.8656,
+      "step": 2094080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5704963367167575e-05,
+      "loss": 3.8617,
+      "step": 2094592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5696577419657055e-05,
+      "loss": 3.8696,
+      "step": 2095104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5688191472146535e-05,
+      "loss": 3.8522,
+      "step": 2095616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5679805524636015e-05,
+      "loss": 3.8631,
+      "step": 2096128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5671419577125494e-05,
+      "loss": 3.8749,
+      "step": 2096640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5663050008418707e-05,
+      "loss": 3.8675,
+      "step": 2097152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5654664060908187e-05,
+      "loss": 3.879,
+      "step": 2097664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5646278113397667e-05,
+      "loss": 3.8704,
+      "step": 2098176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5637892165887147e-05,
+      "loss": 3.8863,
+      "step": 2098688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5629522597180356e-05,
+      "loss": 3.8593,
+      "step": 2099200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5621136649669836e-05,
+      "loss": 3.877,
+      "step": 2099712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.561275070215932e-05,
+      "loss": 3.8734,
+      "step": 2100224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5604364754648796e-05,
+      "loss": 3.8554,
+      "step": 2100736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5595978807138276e-05,
+      "loss": 3.8751,
+      "step": 2101248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5587592859627756e-05,
+      "loss": 3.8728,
+      "step": 2101760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5579206912117236e-05,
+      "loss": 3.8725,
+      "step": 2102272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5570837343410448e-05,
+      "loss": 3.8788,
+      "step": 2102784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5562451395899928e-05,
+      "loss": 3.8619,
+      "step": 2103296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5554065448389408e-05,
+      "loss": 3.8518,
+      "step": 2103808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5545679500878888e-05,
+      "loss": 3.8609,
+      "step": 2104320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5537293553368368e-05,
+      "loss": 3.8728,
+      "step": 2104832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5528907605857848e-05,
+      "loss": 3.8664,
+      "step": 2105344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5520521658347328e-05,
+      "loss": 3.8641,
+      "step": 2105856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5512135710836804e-05,
+      "loss": 3.8628,
+      "step": 2106368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5503749763326288e-05,
+      "loss": 3.8686,
+      "step": 2106880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.549539657342323e-05,
+      "loss": 3.8663,
+      "step": 2107392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.548701062591271e-05,
+      "loss": 3.8665,
+      "step": 2107904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.547862467840219e-05,
+      "loss": 3.8771,
+      "step": 2108416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.547023873089167e-05,
+      "loss": 3.8813,
+      "step": 2108928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.546185278338115e-05,
+      "loss": 3.874,
+      "step": 2109440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.545346683587063e-05,
+      "loss": 3.8659,
+      "step": 2109952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.544508088836011e-05,
+      "loss": 3.8656,
+      "step": 2110464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.543671131965332e-05,
+      "loss": 3.8643,
+      "step": 2110976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5428325372142798e-05,
+      "loss": 3.8772,
+      "step": 2111488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5419939424632278e-05,
+      "loss": 3.8702,
+      "step": 2112000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5411553477121758e-05,
+      "loss": 3.8686,
+      "step": 2112512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.540316752961124e-05,
+      "loss": 3.8687,
+      "step": 2113024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.539479796090445e-05,
+      "loss": 3.8711,
+      "step": 2113536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.538641201339393e-05,
+      "loss": 3.8679,
+      "step": 2114048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.537802606588341e-05,
+      "loss": 3.8666,
+      "step": 2114560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.536964011837289e-05,
+      "loss": 3.8586,
+      "step": 2115072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.536125417086237e-05,
+      "loss": 3.8586,
+      "step": 2115584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.535286822335185e-05,
+      "loss": 3.868,
+      "step": 2116096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5344482275841334e-05,
+      "loss": 3.858,
+      "step": 2116608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.533609632833081e-05,
+      "loss": 3.8678,
+      "step": 2117120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5327726759624023e-05,
+      "loss": 3.8661,
+      "step": 2117632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5319340812113503e-05,
+      "loss": 3.8637,
+      "step": 2118144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.531095486460298e-05,
+      "loss": 3.8591,
+      "step": 2118656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.530256891709246e-05,
+      "loss": 3.8789,
+      "step": 2119168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5294182969581943e-05,
+      "loss": 3.8819,
+      "step": 2119680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5285813400875155e-05,
+      "loss": 3.8674,
+      "step": 2120192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.527742745336463e-05,
+      "loss": 3.8721,
+      "step": 2120704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.526904150585411e-05,
+      "loss": 3.8672,
+      "step": 2121216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.526065555834359e-05,
+      "loss": 3.8662,
+      "step": 2121728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5252269610833073e-05,
+      "loss": 3.8695,
+      "step": 2122240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5243900042126282e-05,
+      "loss": 3.8604,
+      "step": 2122752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5235514094615764e-05,
+      "loss": 3.8651,
+      "step": 2123264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5227128147105244e-05,
+      "loss": 3.8707,
+      "step": 2123776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5218742199594724e-05,
+      "loss": 3.872,
+      "step": 2124288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5210356252084204e-05,
+      "loss": 3.8666,
+      "step": 2124800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5201970304573684e-05,
+      "loss": 3.8687,
+      "step": 2125312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5193584357063162e-05,
+      "loss": 3.8692,
+      "step": 2125824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5185198409552642e-05,
+      "loss": 3.8619,
+      "step": 2126336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5176812462042122e-05,
+      "loss": 3.8718,
+      "step": 2126848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5168442893335336e-05,
+      "loss": 3.8758,
+      "step": 2127360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5160056945824813e-05,
+      "loss": 3.8782,
+      "step": 2127872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5151670998314294e-05,
+      "loss": 3.8602,
+      "step": 2128384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5143285050803774e-05,
+      "loss": 3.8824,
+      "step": 2128896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5134915482096987e-05,
+      "loss": 3.869,
+      "step": 2129408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5126529534586465e-05,
+      "loss": 3.877,
+      "step": 2129920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5118143587075945e-05,
+      "loss": 3.8737,
+      "step": 2130432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5109757639565425e-05,
+      "loss": 3.868,
+      "step": 2130944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5101388070858636e-05,
+      "loss": 3.8614,
+      "step": 2131456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5093002123348116e-05,
+      "loss": 3.882,
+      "step": 2131968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5084616175837596e-05,
+      "loss": 3.8553,
+      "step": 2132480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5076230228327076e-05,
+      "loss": 3.8717,
+      "step": 2132992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5067860659620286e-05,
+      "loss": 3.8718,
+      "step": 2133504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5059474712109766e-05,
+      "loss": 3.8716,
+      "step": 2134016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.505110514340298e-05,
+      "loss": 3.8672,
+      "step": 2134528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.504271919589246e-05,
+      "loss": 3.8676,
+      "step": 2135040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5034333248381937e-05,
+      "loss": 3.8676,
+      "step": 2135552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5025947300871419e-05,
+      "loss": 3.8801,
+      "step": 2136064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5017561353360899e-05,
+      "loss": 3.8705,
+      "step": 2136576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9788289070129395,
+      "eval_runtime": 329.3101,
+      "eval_samples_per_second": 1158.759,
+      "eval_steps_per_second": 36.212,
+      "step": 2136948
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5009175405850379e-05,
+      "loss": 3.8522,
+      "step": 2137088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5000789458339859e-05,
+      "loss": 3.863,
+      "step": 2137600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4992403510829338e-05,
+      "loss": 3.8669,
+      "step": 2138112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4984017563318817e-05,
+      "loss": 3.873,
+      "step": 2138624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4975631615808297e-05,
+      "loss": 3.8784,
+      "step": 2139136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4967245668297777e-05,
+      "loss": 3.8629,
+      "step": 2139648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4958859720787258e-05,
+      "loss": 3.8723,
+      "step": 2140160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4950473773276738e-05,
+      "loss": 3.8674,
+      "step": 2140672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4942087825766218e-05,
+      "loss": 3.8598,
+      "step": 2141184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4933701878255698e-05,
+      "loss": 3.8567,
+      "step": 2141696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4925315930745176e-05,
+      "loss": 3.8777,
+      "step": 2142208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4916929983234656e-05,
+      "loss": 3.8803,
+      "step": 2142720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4908560414527869e-05,
+      "loss": 3.8752,
+      "step": 2143232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.490017446701735e-05,
+      "loss": 3.8538,
+      "step": 2143744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4891788519506827e-05,
+      "loss": 3.8709,
+      "step": 2144256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4883402571996307e-05,
+      "loss": 3.8552,
+      "step": 2144768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4875016624485789e-05,
+      "loss": 3.8711,
+      "step": 2145280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4866630676975269e-05,
+      "loss": 3.8654,
+      "step": 2145792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4858244729464749e-05,
+      "loss": 3.8647,
+      "step": 2146304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4849858781954229e-05,
+      "loss": 3.872,
+      "step": 2146816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4841472834443707e-05,
+      "loss": 3.8769,
+      "step": 2147328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4833086886933187e-05,
+      "loss": 3.8674,
+      "step": 2147840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4824700939422667e-05,
+      "loss": 3.8702,
+      "step": 2148352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4816314991912147e-05,
+      "loss": 3.8658,
+      "step": 2148864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4807945423205357e-05,
+      "loss": 3.872,
+      "step": 2149376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4799559475694837e-05,
+      "loss": 3.8731,
+      "step": 2149888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4791173528184319e-05,
+      "loss": 3.8585,
+      "step": 2150400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4782803959477528e-05,
+      "loss": 3.8621,
+      "step": 2150912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4774418011967008e-05,
+      "loss": 3.868,
+      "step": 2151424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.476603206445649e-05,
+      "loss": 3.8644,
+      "step": 2151936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.475764611694597e-05,
+      "loss": 3.8634,
+      "step": 2152448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.474926016943545e-05,
+      "loss": 3.8686,
+      "step": 2152960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.474089060072866e-05,
+      "loss": 3.8719,
+      "step": 2153472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.473250465321814e-05,
+      "loss": 3.8748,
+      "step": 2153984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.472411870570762e-05,
+      "loss": 3.8655,
+      "step": 2154496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4715749137000831e-05,
+      "loss": 3.8677,
+      "step": 2155008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4707363189490311e-05,
+      "loss": 3.8766,
+      "step": 2155520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4698977241979791e-05,
+      "loss": 3.8611,
+      "step": 2156032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4690591294469273e-05,
+      "loss": 3.8593,
+      "step": 2156544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4682205346958753e-05,
+      "loss": 3.8617,
+      "step": 2157056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4673819399448233e-05,
+      "loss": 3.8588,
+      "step": 2157568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4665433451937711e-05,
+      "loss": 3.8664,
+      "step": 2158080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4657047504427191e-05,
+      "loss": 3.8602,
+      "step": 2158592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.464866155691667e-05,
+      "loss": 3.868,
+      "step": 2159104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.464027560940615e-05,
+      "loss": 3.8707,
+      "step": 2159616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.463188966189563e-05,
+      "loss": 3.8738,
+      "step": 2160128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4623503714385112e-05,
+      "loss": 3.8679,
+      "step": 2160640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4615134145678321e-05,
+      "loss": 3.8689,
+      "step": 2161152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4606748198167801e-05,
+      "loss": 3.8474,
+      "step": 2161664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4598362250657283e-05,
+      "loss": 3.8772,
+      "step": 2162176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4589976303146763e-05,
+      "loss": 3.8624,
+      "step": 2162688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4581590355636243e-05,
+      "loss": 3.8701,
+      "step": 2163200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4573220786929454e-05,
+      "loss": 3.8641,
+      "step": 2163712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4564834839418934e-05,
+      "loss": 3.8633,
+      "step": 2164224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4556448891908414e-05,
+      "loss": 3.8563,
+      "step": 2164736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4548079323201624e-05,
+      "loss": 3.8565,
+      "step": 2165248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4539693375691104e-05,
+      "loss": 3.864,
+      "step": 2165760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4531307428180584e-05,
+      "loss": 3.8547,
+      "step": 2166272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4522921480670066e-05,
+      "loss": 3.8746,
+      "step": 2166784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4514535533159543e-05,
+      "loss": 3.8668,
+      "step": 2167296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4506149585649023e-05,
+      "loss": 3.8525,
+      "step": 2167808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4497763638138504e-05,
+      "loss": 3.8778,
+      "step": 2168320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4489377690627984e-05,
+      "loss": 3.8542,
+      "step": 2168832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4480991743117464e-05,
+      "loss": 3.8529,
+      "step": 2169344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4472622174410675e-05,
+      "loss": 3.8691,
+      "step": 2169856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4464236226900155e-05,
+      "loss": 3.8608,
+      "step": 2170368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4455850279389635e-05,
+      "loss": 3.8561,
+      "step": 2170880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4447464331879115e-05,
+      "loss": 3.8655,
+      "step": 2171392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4439078384368596e-05,
+      "loss": 3.8522,
+      "step": 2171904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4430708815661806e-05,
+      "loss": 3.8545,
+      "step": 2172416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4422339246955016e-05,
+      "loss": 3.8739,
+      "step": 2172928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4413953299444496e-05,
+      "loss": 3.8657,
+      "step": 2173440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4405567351933976e-05,
+      "loss": 3.8705,
+      "step": 2173952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4397181404423458e-05,
+      "loss": 3.8684,
+      "step": 2174464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4388795456912938e-05,
+      "loss": 3.8809,
+      "step": 2174976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4380425888206147e-05,
+      "loss": 3.8548,
+      "step": 2175488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4372039940695629e-05,
+      "loss": 3.875,
+      "step": 2176000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4363653993185109e-05,
+      "loss": 3.8656,
+      "step": 2176512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4355268045674589e-05,
+      "loss": 3.8571,
+      "step": 2177024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4346882098164068e-05,
+      "loss": 3.8679,
+      "step": 2177536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4338496150653547e-05,
+      "loss": 3.8716,
+      "step": 2178048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4330110203143027e-05,
+      "loss": 3.8681,
+      "step": 2178560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4321724255632507e-05,
+      "loss": 3.8791,
+      "step": 2179072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.431335468692572e-05,
+      "loss": 3.8556,
+      "step": 2179584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4304968739415197e-05,
+      "loss": 3.8489,
+      "step": 2180096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4296582791904677e-05,
+      "loss": 3.8553,
+      "step": 2180608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4288196844394159e-05,
+      "loss": 3.869,
+      "step": 2181120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.42798436544911e-05,
+      "loss": 3.8661,
+      "step": 2181632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4271474085784311e-05,
+      "loss": 3.859,
+      "step": 2182144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4263088138273791e-05,
+      "loss": 3.8615,
+      "step": 2182656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4254702190763273e-05,
+      "loss": 3.866,
+      "step": 2183168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4246316243252753e-05,
+      "loss": 3.8587,
+      "step": 2183680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4237930295742233e-05,
+      "loss": 3.862,
+      "step": 2184192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4229544348231713e-05,
+      "loss": 3.8719,
+      "step": 2184704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4221158400721193e-05,
+      "loss": 3.876,
+      "step": 2185216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4212772453210671e-05,
+      "loss": 3.8699,
+      "step": 2185728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4204386505700151e-05,
+      "loss": 3.8579,
+      "step": 2186240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4196000558189631e-05,
+      "loss": 3.8621,
+      "step": 2186752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4187614610679113e-05,
+      "loss": 3.8589,
+      "step": 2187264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4179228663168593e-05,
+      "loss": 3.8748,
+      "step": 2187776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4170859094461802e-05,
+      "loss": 3.87,
+      "step": 2188288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4162473146951283e-05,
+      "loss": 3.8631,
+      "step": 2188800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4154087199440763e-05,
+      "loss": 3.8664,
+      "step": 2189312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4145701251930243e-05,
+      "loss": 3.8627,
+      "step": 2189824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4137331683223454e-05,
+      "loss": 3.8659,
+      "step": 2190336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4128962114516667e-05,
+      "loss": 3.8605,
+      "step": 2190848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4120576167006145e-05,
+      "loss": 3.8583,
+      "step": 2191360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4112190219495625e-05,
+      "loss": 3.8521,
+      "step": 2191872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4103804271985105e-05,
+      "loss": 3.8616,
+      "step": 2192384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4095418324474585e-05,
+      "loss": 3.8603,
+      "step": 2192896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4087032376964066e-05,
+      "loss": 3.8579,
+      "step": 2193408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4078646429453546e-05,
+      "loss": 3.8604,
+      "step": 2193920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4070260481943023e-05,
+      "loss": 3.8612,
+      "step": 2194432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4061890913236237e-05,
+      "loss": 3.8569,
+      "step": 2194944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4053504965725717e-05,
+      "loss": 3.8715,
+      "step": 2195456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4045119018215197e-05,
+      "loss": 3.878,
+      "step": 2195968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4036733070704675e-05,
+      "loss": 3.8673,
+      "step": 2196480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4028347123194155e-05,
+      "loss": 3.8672,
+      "step": 2196992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4019977554487368e-05,
+      "loss": 3.8636,
+      "step": 2197504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4011591606976848e-05,
+      "loss": 3.8624,
+      "step": 2198016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4003205659466326e-05,
+      "loss": 3.8691,
+      "step": 2198528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3994836090759538e-05,
+      "loss": 3.8509,
+      "step": 2199040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.398645014324902e-05,
+      "loss": 3.8633,
+      "step": 2199552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3978064195738497e-05,
+      "loss": 3.8637,
+      "step": 2200064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3969678248227977e-05,
+      "loss": 3.8695,
+      "step": 2200576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3961292300717456e-05,
+      "loss": 3.8596,
+      "step": 2201088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3952906353206938e-05,
+      "loss": 3.8677,
+      "step": 2201600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3944520405696418e-05,
+      "loss": 3.8636,
+      "step": 2202112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3936134458185898e-05,
+      "loss": 3.8566,
+      "step": 2202624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3927764889479109e-05,
+      "loss": 3.8705,
+      "step": 2203136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3919378941968589e-05,
+      "loss": 3.8684,
+      "step": 2203648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3910992994458069e-05,
+      "loss": 3.8747,
+      "step": 2204160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3902607046947549e-05,
+      "loss": 3.8584,
+      "step": 2204672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.389422109943703e-05,
+      "loss": 3.8772,
+      "step": 2205184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3885835151926507e-05,
+      "loss": 3.8666,
+      "step": 2205696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3877465583219721e-05,
+      "loss": 3.8705,
+      "step": 2206208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3869079635709201e-05,
+      "loss": 3.8717,
+      "step": 2206720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3860693688198678e-05,
+      "loss": 3.8633,
+      "step": 2207232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.385230774068816e-05,
+      "loss": 3.8571,
+      "step": 2207744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3843938171981372e-05,
+      "loss": 3.8765,
+      "step": 2208256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3835552224470852e-05,
+      "loss": 3.8534,
+      "step": 2208768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.382716627696033e-05,
+      "loss": 3.8625,
+      "step": 2209280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3818796708253542e-05,
+      "loss": 3.8684,
+      "step": 2209792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3810410760743022e-05,
+      "loss": 3.8707,
+      "step": 2210304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3802024813232502e-05,
+      "loss": 3.8639,
+      "step": 2210816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.379363886572198e-05,
+      "loss": 3.8631,
+      "step": 2211328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.378525291821146e-05,
+      "loss": 3.8659,
+      "step": 2211840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.377686697070094e-05,
+      "loss": 3.8713,
+      "step": 2212352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3768481023190422e-05,
+      "loss": 3.8694,
+      "step": 2212864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.977329730987549,
+      "eval_runtime": 307.3862,
+      "eval_samples_per_second": 1241.406,
+      "eval_steps_per_second": 38.795,
+      "step": 2213268
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 9.140026793693548e+17,
+  "trial_name": null,
+  "trial_params": null
+}