diff --git "a/checkpoint-686880/trainer_state.json" "b/checkpoint-686880/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-686880/trainer_state.json"
@@ -0,0 +1,8143 @@
+{
+  "best_metric": 4.089617729187012,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-c-command/lstm/1/checkpoints/checkpoint-686880",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 686880,
+  "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": 10.8194,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5501,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0562,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9924,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9445,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8914,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7191,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6279,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5288,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.462,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3894,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3229,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2581,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1826,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1381,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0784,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0419,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9957,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9488,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.912,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.874,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.837,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8066,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7679,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7509,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7201,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6907,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6686,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6417,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975684027980238e-05,
+      "loss": 5.6247,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974845433229186e-05,
+      "loss": 5.6047,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5844,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5627,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5411,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5372,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5275,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.4998,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.4604,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4654,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4356,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4355,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4196,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4084,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.384,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3881,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3761,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3727,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.357,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3364,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3297,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3319,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3092,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3015,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.2876,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.284,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2673,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2818,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2458,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2509,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2428,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2222,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2247,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2126,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.1955,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.1887,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2059,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1805,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1722,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1541,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1626,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1683,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1518,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1462,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1518,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1414,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1299,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1089,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.116,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.098,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0973,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0911,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0873,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0955,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0856,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0583,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0598,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0627,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0587,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0486,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0433,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0367,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0228,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0289,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0223,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.012,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0188,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.9993,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9971,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9859,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.983,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 4.9908,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9747,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9708,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9656,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.964,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9471,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.9365,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9436,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9494,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.9396,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.9366,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.9222,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9287,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.9155,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9188,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.914,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9113,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9025,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.899,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.9028,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.901,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.889,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8911,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8816,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8833,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.8635,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8586,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8633,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8646,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8555,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.8669,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.8577,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8415,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8467,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8514,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8423,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.839,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.8298,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884314871364151e-05,
+      "loss": 4.8331,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.8361,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8186,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8113,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8184,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8102,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.8063,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.8098,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.8018,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.7994,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7982,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7924,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.759008407592773,
+      "eval_runtime": 288.2778,
+      "eval_samples_per_second": 1323.692,
+      "eval_steps_per_second": 41.366,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.7851,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7888,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7948,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7835,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7825,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7765,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7791,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7589,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7814,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667092952331785e-05,
+      "loss": 4.7678,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7744,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7718,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7523,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7555,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7484,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7433,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7517,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7372,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7357,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7545,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7369,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7327,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.7259,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.7404,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7188,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7143,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.7154,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.7184,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7083,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.7104,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.7109,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.7105,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.6976,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.7107,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.7124,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.7039,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.6774,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.6863,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.6824,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.6896,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.6799,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.6839,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.6711,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838206900979649e-05,
+      "loss": 4.686,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6803,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.6805,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6798,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6597,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6686,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.6727,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323416513634046e-05,
+      "loss": 4.6669,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315030566123526e-05,
+      "loss": 4.6642,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306644618613006e-05,
+      "loss": 4.6513,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.659,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.6427,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.6724,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.6443,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.6551,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825637806996108e-05,
+      "loss": 4.6521,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824799212245056e-05,
+      "loss": 4.6297,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.6459,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.6388,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.6254,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.6252,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820606238489796e-05,
+      "loss": 4.6435,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819767643738744e-05,
+      "loss": 4.6304,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.6236,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6174,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.6307,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.6362,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155779457442305e-05,
+      "loss": 4.6292,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.6214,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139007562421265e-05,
+      "loss": 4.6371,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130621614910745e-05,
+      "loss": 4.6302,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122235667400224e-05,
+      "loss": 4.6258,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.6075,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.6194,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809707782486866e-05,
+      "loss": 4.6119,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.6071,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.609,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.6083,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6216,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.6164,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046794897413e-05,
+      "loss": 4.5923,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803840894990248e-05,
+      "loss": 4.597,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803002300239196e-05,
+      "loss": 4.6016,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.6017,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.6011,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800488153866413e-05,
+      "loss": 4.5991,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.594,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.5846,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.5951,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797133774862205e-05,
+      "loss": 4.5879,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796295180111153e-05,
+      "loss": 4.5879,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.5957,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5843,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.5777,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792942438987319e-05,
+      "loss": 4.5784,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.5807,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.5853,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5718,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.5705,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5721,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.572,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5606,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862369567396485e-05,
+      "loss": 4.5475,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853983619885965e-05,
+      "loss": 4.5643,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845597672375445e-05,
+      "loss": 4.5659,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837211724864925e-05,
+      "loss": 4.5633,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.562,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.5553,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.5612,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.5501,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.5531,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.5574,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.5576,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777014052358449e-05,
+      "loss": 4.5491,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776175457607397e-05,
+      "loss": 4.5496,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.5569,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5557,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.5407,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.5521,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.5397,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.5524,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77030857011078e-05,
+      "loss": 4.5299,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.5295,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768631380608676e-05,
+      "loss": 4.5341,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.5408,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766954191106572e-05,
+      "loss": 4.5298,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76611559635552e-05,
+      "loss": 4.5479,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5368,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.5288,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.5309,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762762855231685e-05,
+      "loss": 4.5405,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761925898361006e-05,
+      "loss": 4.532,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.5324,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760248708858902e-05,
+      "loss": 4.5242,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941011410785e-05,
+      "loss": 4.527,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758573157237171e-05,
+      "loss": 4.5324,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757734562486119e-05,
+      "loss": 4.5202,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756895967735067e-05,
+      "loss": 4.5188,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756057372984015e-05,
+      "loss": 4.5248,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755220416113336e-05,
+      "loss": 4.5151,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754381821362284e-05,
+      "loss": 4.5153,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753543226611232e-05,
+      "loss": 4.5231,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75270463186018e-05,
+      "loss": 4.5133,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518676749895016e-05,
+      "loss": 4.513,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510307181188225e-05,
+      "loss": 4.5124,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501921233677705e-05,
+      "loss": 4.5076,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.479434013366699,
+      "eval_runtime": 280.6715,
+      "eval_samples_per_second": 1359.564,
+      "eval_steps_per_second": 42.487,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493535286167185e-05,
+      "loss": 4.5016,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485149338656665e-05,
+      "loss": 4.5081,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476763391146145e-05,
+      "loss": 4.5194,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468377443635625e-05,
+      "loss": 4.5095,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459991496125105e-05,
+      "loss": 4.5078,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451605548614585e-05,
+      "loss": 4.5018,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443219601104065e-05,
+      "loss": 4.5093,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434833653593545e-05,
+      "loss": 4.4899,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426447706083025e-05,
+      "loss": 4.5133,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418061758572505e-05,
+      "loss": 4.509,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409675811061985e-05,
+      "loss": 4.5044,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401289863551465e-05,
+      "loss": 4.5135,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739292029484468e-05,
+      "loss": 4.4894,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384534347334154e-05,
+      "loss": 4.495,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376148399823634e-05,
+      "loss": 4.5011,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367762452313114e-05,
+      "loss": 4.4827,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359376504802594e-05,
+      "loss": 4.4934,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350990557292074e-05,
+      "loss": 4.4848,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342604609781554e-05,
+      "loss": 4.4851,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334218662271034e-05,
+      "loss": 4.5079,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325832714760514e-05,
+      "loss": 4.4867,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317446767249994e-05,
+      "loss": 4.4907,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309060819739474e-05,
+      "loss": 4.4781,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730069125103269e-05,
+      "loss": 4.4981,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729230530352217e-05,
+      "loss": 4.4732,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728391935601165e-05,
+      "loss": 4.4749,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727553340850113e-05,
+      "loss": 4.473,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726716383979434e-05,
+      "loss": 4.4831,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725879427108755e-05,
+      "loss": 4.4693,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725040832357703e-05,
+      "loss": 4.4723,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724202237606651e-05,
+      "loss": 4.4793,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723363642855599e-05,
+      "loss": 4.4804,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722525048104547e-05,
+      "loss": 4.4621,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721686453353495e-05,
+      "loss": 4.4798,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720847858602443e-05,
+      "loss": 4.4817,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720009263851391e-05,
+      "loss": 4.4727,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719172306980712e-05,
+      "loss": 4.4514,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718335350110033e-05,
+      "loss": 4.4645,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717496755358981e-05,
+      "loss": 4.4524,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716658160607929e-05,
+      "loss": 4.4693,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715819565856877e-05,
+      "loss": 4.4608,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714980971105825e-05,
+      "loss": 4.4641,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714142376354773e-05,
+      "loss": 4.453,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713303781603721e-05,
+      "loss": 4.4622,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712466824733042e-05,
+      "loss": 4.4634,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71162822998199e-05,
+      "loss": 4.4646,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710789635230938e-05,
+      "loss": 4.4655,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709951040479886e-05,
+      "loss": 4.4437,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709114083609208e-05,
+      "loss": 4.4509,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708275488858156e-05,
+      "loss": 4.4647,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074368941071037e-05,
+      "loss": 4.4507,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065982993560517e-05,
+      "loss": 4.4523,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705759704604999e-05,
+      "loss": 4.4443,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704921109853947e-05,
+      "loss": 4.4504,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704082515102895e-05,
+      "loss": 4.4383,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703245558232216e-05,
+      "loss": 4.4591,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702406963481164e-05,
+      "loss": 4.4444,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701568368730112e-05,
+      "loss": 4.4457,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70072977397906e-05,
+      "loss": 4.4554,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699891179228008e-05,
+      "loss": 4.4262,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699052584476956e-05,
+      "loss": 4.4443,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982139897259045e-05,
+      "loss": 4.4363,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973753949748525e-05,
+      "loss": 4.4295,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965384381041734e-05,
+      "loss": 4.4263,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956998433531214e-05,
+      "loss": 4.4402,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694862886482442e-05,
+      "loss": 4.4372,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69402429173139e-05,
+      "loss": 4.4264,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693185696980338e-05,
+      "loss": 4.4198,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692347102229286e-05,
+      "loss": 4.4383,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691508507478234e-05,
+      "loss": 4.4466,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690669912727182e-05,
+      "loss": 4.4343,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68983131797613e-05,
+      "loss": 4.4339,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688992723225078e-05,
+      "loss": 4.4439,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881557663544e-05,
+      "loss": 4.4397,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687317171603348e-05,
+      "loss": 4.4397,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686478576852296e-05,
+      "loss": 4.4159,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685639982101244e-05,
+      "loss": 4.4339,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684803025230565e-05,
+      "loss": 4.433,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683964430479513e-05,
+      "loss": 4.4181,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683127473608834e-05,
+      "loss": 4.4268,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682288878857782e-05,
+      "loss": 4.4253,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68145028410673e-05,
+      "loss": 4.4393,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680611689355678e-05,
+      "loss": 4.4308,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797747324849986e-05,
+      "loss": 4.412,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789361377339466e-05,
+      "loss": 4.4169,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678097542982895e-05,
+      "loss": 4.42,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677258948231843e-05,
+      "loss": 4.4252,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676420353480791e-05,
+      "loss": 4.4226,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675581758729739e-05,
+      "loss": 4.4194,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674743163978687e-05,
+      "loss": 4.4173,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739045692276346e-05,
+      "loss": 4.4075,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673067612356956e-05,
+      "loss": 4.4206,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672229017605904e-05,
+      "loss": 4.4104,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671392060735225e-05,
+      "loss": 4.4169,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670553465984173e-05,
+      "loss": 4.4223,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669714871233121e-05,
+      "loss": 4.4149,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668876276482069e-05,
+      "loss": 4.4066,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668037681731017e-05,
+      "loss": 4.4083,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667199086979965e-05,
+      "loss": 4.4098,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666360492228913e-05,
+      "loss": 4.4142,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665521897477861e-05,
+      "loss": 4.405,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664683302726809e-05,
+      "loss": 4.4021,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638479837365035e-05,
+      "loss": 4.4007,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630093889854515e-05,
+      "loss": 4.407,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621707942343995e-05,
+      "loss": 4.396,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661332199483347e-05,
+      "loss": 4.3861,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660493604732295e-05,
+      "loss": 4.397,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659655009981243e-05,
+      "loss": 4.4032,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658816415230191e-05,
+      "loss": 4.395,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657977820479139e-05,
+      "loss": 4.403,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571408636084604e-05,
+      "loss": 4.3903,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563022688574084e-05,
+      "loss": 4.4004,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655465311986729e-05,
+      "loss": 4.3903,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654626717235677e-05,
+      "loss": 4.389,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653788122484625e-05,
+      "loss": 4.4013,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652949527733573e-05,
+      "loss": 4.3996,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652110932982521e-05,
+      "loss": 4.3921,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651272338231469e-05,
+      "loss": 4.3935,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650433743480417e-05,
+      "loss": 4.3933,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.3973,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648758191858686e-05,
+      "loss": 4.3861,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647919597107634e-05,
+      "loss": 4.3978,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647081002356582e-05,
+      "loss": 4.3855,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646242407605531e-05,
+      "loss": 4.397,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645405450734852e-05,
+      "loss": 4.3732,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445668559838e-05,
+      "loss": 4.3828,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643728261232748e-05,
+      "loss": 4.3783,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642889666481696e-05,
+      "loss": 4.3872,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642051071730644e-05,
+      "loss": 4.3794,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641212476979592e-05,
+      "loss": 4.3982,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403755201089126e-05,
+      "loss": 4.3862,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395369253578606e-05,
+      "loss": 4.381,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6386983306068086e-05,
+      "loss": 4.377,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378613737361295e-05,
+      "loss": 4.3885,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370227789850775e-05,
+      "loss": 4.3863,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636184184234026e-05,
+      "loss": 4.3881,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635345589482974e-05,
+      "loss": 4.3782,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634506994731922e-05,
+      "loss": 4.378,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63366839998087e-05,
+      "loss": 4.3879,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632829805229818e-05,
+      "loss": 4.3757,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319912104787655e-05,
+      "loss": 4.3737,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311526157277135e-05,
+      "loss": 4.3813,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303140209766615e-05,
+      "loss": 4.373,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294754262256095e-05,
+      "loss": 4.3693,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286368314745575e-05,
+      "loss": 4.3756,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277998746038784e-05,
+      "loss": 4.3769,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269629177332e-05,
+      "loss": 4.368,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626124322982148e-05,
+      "loss": 4.368,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625285728231096e-05,
+      "loss": 4.3635,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.345666408538818,
+      "eval_runtime": 303.4225,
+      "eval_samples_per_second": 1257.623,
+      "eval_steps_per_second": 39.302,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624447133480044e-05,
+      "loss": 4.3721,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623608538728992e-05,
+      "loss": 4.3656,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62276994397794e-05,
+      "loss": 4.3799,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621931349226888e-05,
+      "loss": 4.3686,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621092754475836e-05,
+      "loss": 4.3736,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620255797605157e-05,
+      "loss": 4.3624,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619417202854105e-05,
+      "loss": 4.3716,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618578608103053e-05,
+      "loss": 4.3536,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617740013352001e-05,
+      "loss": 4.3765,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616901418600949e-05,
+      "loss": 4.3728,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160660996106434e-05,
+      "loss": 4.3708,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152275048595913e-05,
+      "loss": 4.379,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614390547988913e-05,
+      "loss": 4.3525,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61355195323786e-05,
+      "loss": 4.3608,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612713358486808e-05,
+      "loss": 4.3668,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611874763735756e-05,
+      "loss": 4.3537,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611036168984704e-05,
+      "loss": 4.3565,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610197574233652e-05,
+      "loss": 4.3514,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093589794826e-05,
+      "loss": 4.3535,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608520384731548e-05,
+      "loss": 4.3788,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607681789980496e-05,
+      "loss": 4.3542,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606843195229444e-05,
+      "loss": 4.3563,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606004600478392e-05,
+      "loss": 4.3492,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60516600572734e-05,
+      "loss": 4.3653,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604329048856662e-05,
+      "loss": 4.3465,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60349045410561e-05,
+      "loss": 4.3435,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602651859354558e-05,
+      "loss": 4.3454,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601813264603506e-05,
+      "loss": 4.3551,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600976307732827e-05,
+      "loss": 4.3403,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600137712981775e-05,
+      "loss": 4.3478,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599299118230723e-05,
+      "loss": 4.348,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598460523479671e-05,
+      "loss": 4.3547,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976235666089916e-05,
+      "loss": 4.343,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967849718579396e-05,
+      "loss": 4.3495,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959463771068876e-05,
+      "loss": 4.3556,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951077823558356e-05,
+      "loss": 4.3483,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59427246336553e-05,
+      "loss": 4.3277,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593433868614478e-05,
+      "loss": 4.3438,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592595273863426e-05,
+      "loss": 4.3274,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591756679112374e-05,
+      "loss": 4.3466,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590918084361322e-05,
+      "loss": 4.3346,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59007948961027e-05,
+      "loss": 4.3452,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589240894859218e-05,
+      "loss": 4.335,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588402300108166e-05,
+      "loss": 4.3392,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587565343237487e-05,
+      "loss": 4.3447,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586726748486435e-05,
+      "loss": 4.3393,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585888153735383e-05,
+      "loss": 4.3437,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585049558984331e-05,
+      "loss": 4.329,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584210964233279e-05,
+      "loss": 4.3281,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833740073626005e-05,
+      "loss": 4.3442,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825354126115485e-05,
+      "loss": 4.3324,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816968178604965e-05,
+      "loss": 4.3307,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580858223109444e-05,
+      "loss": 4.3283,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580022904119138e-05,
+      "loss": 4.334,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579184309368086e-05,
+      "loss": 4.3195,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578345714617034e-05,
+      "loss": 4.3399,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577507119865982e-05,
+      "loss": 4.3268,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57666852511493e-05,
+      "loss": 4.3318,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575829930363878e-05,
+      "loss": 4.336,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574991335612826e-05,
+      "loss": 4.3176,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574152740861774e-05,
+      "loss": 4.3222,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573315783991096e-05,
+      "loss": 4.3255,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572477189240044e-05,
+      "loss": 4.314,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571638594488991e-05,
+      "loss": 4.3135,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570799999737939e-05,
+      "loss": 4.3238,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569963042867261e-05,
+      "loss": 4.3268,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569124448116208e-05,
+      "loss": 4.3124,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568285853365156e-05,
+      "loss": 4.3082,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567447258614104e-05,
+      "loss": 4.3231,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566610301743426e-05,
+      "loss": 4.3338,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565771706992373e-05,
+      "loss": 4.3206,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564933112241322e-05,
+      "loss": 4.3249,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564096155370643e-05,
+      "loss": 4.3299,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563257560619591e-05,
+      "loss": 4.3305,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624189658685386e-05,
+      "loss": 4.3312,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615803711174866e-05,
+      "loss": 4.305,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607417763664346e-05,
+      "loss": 4.3226,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599031816153826e-05,
+      "loss": 4.3262,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590645868643306e-05,
+      "loss": 4.3062,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582259921132786e-05,
+      "loss": 4.3219,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557390673122973e-05,
+      "loss": 4.3144,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565520783719204e-05,
+      "loss": 4.3288,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557134836208684e-05,
+      "loss": 4.3252,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554874888869817e-05,
+      "loss": 4.3021,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554036294118765e-05,
+      "loss": 4.3133,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553197699367713e-05,
+      "loss": 4.3107,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552359104616661e-05,
+      "loss": 4.3167,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551522147745982e-05,
+      "loss": 4.3179,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55068355299493e-05,
+      "loss": 4.3119,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549844958243878e-05,
+      "loss": 4.3097,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549006363492826e-05,
+      "loss": 4.3034,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548169406622147e-05,
+      "loss": 4.3131,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547330811871095e-05,
+      "loss": 4.3042,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546492217120043e-05,
+      "loss": 4.3133,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545653622368991e-05,
+      "loss": 4.3206,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544815027617939e-05,
+      "loss": 4.3079,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543976432866887e-05,
+      "loss": 4.3032,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431378381158355e-05,
+      "loss": 4.3076,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422992433647835e-05,
+      "loss": 4.3048,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414622864941044e-05,
+      "loss": 4.3111,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406236917430524e-05,
+      "loss": 4.3056,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397850969920004e-05,
+      "loss": 4.3005,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389465022409484e-05,
+      "loss": 4.2956,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538109545370269e-05,
+      "loss": 4.3089,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537270950619217e-05,
+      "loss": 4.2903,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536432355868165e-05,
+      "loss": 4.2872,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535595398997486e-05,
+      "loss": 4.2939,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534756804246434e-05,
+      "loss": 4.2996,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533918209495382e-05,
+      "loss": 4.2973,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533079614744331e-05,
+      "loss": 4.3092,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532241019993279e-05,
+      "loss": 4.2915,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531402425242227e-05,
+      "loss": 4.2925,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530563830491175e-05,
+      "loss": 4.2961,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529725235740122e-05,
+      "loss": 4.2853,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528888278869444e-05,
+      "loss": 4.3079,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528049684118392e-05,
+      "loss": 4.2995,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272127272477126e-05,
+      "loss": 4.2958,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263741324966606e-05,
+      "loss": 4.2963,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255355377456086e-05,
+      "loss": 4.2882,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246969429945566e-05,
+      "loss": 4.3009,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238583482435046e-05,
+      "loss": 4.2885,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230213913728255e-05,
+      "loss": 4.3045,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522182796621774e-05,
+      "loss": 4.2864,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521344201870722e-05,
+      "loss": 4.2944,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5205056071196695e-05,
+      "loss": 4.2829,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196670123686175e-05,
+      "loss": 4.2862,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188284176175655e-05,
+      "loss": 4.2811,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5179898228665135e-05,
+      "loss": 4.2956,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171512281154615e-05,
+      "loss": 4.282,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163142712447824e-05,
+      "loss": 4.3012,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154756764937304e-05,
+      "loss": 4.2945,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146370817426784e-05,
+      "loss": 4.2881,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137984869916264e-05,
+      "loss": 4.2797,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512963168001321e-05,
+      "loss": 4.2947,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512124573250269e-05,
+      "loss": 4.2895,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511285978499217e-05,
+      "loss": 4.2969,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510447383748165e-05,
+      "loss": 4.2843,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509608788997113e-05,
+      "loss": 4.2824,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508770194246061e-05,
+      "loss": 4.2934,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507931599495009e-05,
+      "loss": 4.2885,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507093004743957e-05,
+      "loss": 4.2776,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506254409992905e-05,
+      "loss": 4.288,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505417453122226e-05,
+      "loss": 4.28,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504578858371174e-05,
+      "loss": 4.2773,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503740263620122e-05,
+      "loss": 4.284,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50290166886907e-05,
+      "loss": 4.288,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020647119983914e-05,
+      "loss": 4.2748,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012261172473394e-05,
+      "loss": 4.2755,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003875224962873e-05,
+      "loss": 4.2825,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.262726783752441,
+      "eval_runtime": 300.4001,
+      "eval_samples_per_second": 1270.276,
+      "eval_steps_per_second": 39.697,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995489277452353e-05,
+      "loss": 4.276,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498710332994183e-05,
+      "loss": 4.2757,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497871738243131e-05,
+      "loss": 4.2893,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497033143492079e-05,
+      "loss": 4.2767,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496194548741027e-05,
+      "loss": 4.2895,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495355953989975e-05,
+      "loss": 4.2737,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945173592389226e-05,
+      "loss": 4.2804,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936787644878706e-05,
+      "loss": 4.2664,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928401697368186e-05,
+      "loss": 4.2902,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920015749857666e-05,
+      "loss": 4.2818,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911629802347146e-05,
+      "loss": 4.2802,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490324385483663e-05,
+      "loss": 4.2913,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489489066493358e-05,
+      "loss": 4.2641,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488650471742305e-05,
+      "loss": 4.2771,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487811876991253e-05,
+      "loss": 4.2789,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486973282240201e-05,
+      "loss": 4.2661,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486134687489149e-05,
+      "loss": 4.2697,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485296092738097e-05,
+      "loss": 4.266,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484457497987045e-05,
+      "loss": 4.2704,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483618903235993e-05,
+      "loss": 4.2863,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482780308484941e-05,
+      "loss": 4.269,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481941713733889e-05,
+      "loss": 4.2709,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481103118982837e-05,
+      "loss": 4.2644,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480264524231785e-05,
+      "loss": 4.278,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479425929480733e-05,
+      "loss": 4.2636,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478587334729681e-05,
+      "loss": 4.2599,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47774873997863e-05,
+      "loss": 4.2586,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476910145227577e-05,
+      "loss": 4.2705,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476071550476525e-05,
+      "loss": 4.2541,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752362314862195e-05,
+      "loss": 4.2674,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4743976367351675e-05,
+      "loss": 4.263,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4735590419841155e-05,
+      "loss": 4.2703,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727204472330635e-05,
+      "loss": 4.2614,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718818524820115e-05,
+      "loss": 4.2668,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471043257730959e-05,
+      "loss": 4.2714,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470204662979907e-05,
+      "loss": 4.2664,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469366068228855e-05,
+      "loss": 4.2474,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685274734778035e-05,
+      "loss": 4.2578,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676905166071244e-05,
+      "loss": 4.2471,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668519218560724e-05,
+      "loss": 4.2621,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466014964985394e-05,
+      "loss": 4.2547,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465176370234341e-05,
+      "loss": 4.2628,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464337775483289e-05,
+      "loss": 4.254,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463499180732237e-05,
+      "loss": 4.2601,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462660585981185e-05,
+      "loss": 4.2625,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461821991230133e-05,
+      "loss": 4.258,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460985034359454e-05,
+      "loss": 4.2621,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460146439608402e-05,
+      "loss": 4.2529,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45930784485735e-05,
+      "loss": 4.2421,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458470887986672e-05,
+      "loss": 4.2657,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45763229323562e-05,
+      "loss": 4.2543,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456793698484568e-05,
+      "loss": 4.2512,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455955103733516e-05,
+      "loss": 4.2507,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455116508982464e-05,
+      "loss": 4.2552,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454277914231412e-05,
+      "loss": 4.2395,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45343931948036e-05,
+      "loss": 4.2606,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452600724729308e-05,
+      "loss": 4.2472,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451762129978256e-05,
+      "loss": 4.2523,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450923535227204e-05,
+      "loss": 4.2601,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450084940476152e-05,
+      "loss": 4.244,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492463457251e-05,
+      "loss": 4.2409,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4484093888544207e-05,
+      "loss": 4.2553,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475707941033686e-05,
+      "loss": 4.2328,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446732199352317e-05,
+      "loss": 4.2377,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445893604601265e-05,
+      "loss": 4.2484,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450550098502126e-05,
+      "loss": 4.2498,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442164150991606e-05,
+      "loss": 4.2357,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443379458228482e-05,
+      "loss": 4.2331,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44254086347743e-05,
+      "loss": 4.2447,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4417022687263775e-05,
+      "loss": 4.254,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408636739753255e-05,
+      "loss": 4.2485,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440026717104647e-05,
+      "loss": 4.2465,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439188122353595e-05,
+      "loss": 4.2538,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383495276025424e-05,
+      "loss": 4.2515,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437510932851491e-05,
+      "loss": 4.2614,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436673975980813e-05,
+      "loss": 4.2302,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43583538122976e-05,
+      "loss": 4.247,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434996786478708e-05,
+      "loss": 4.2511,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434158191727656e-05,
+      "loss": 4.2301,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333212348569776e-05,
+      "loss": 4.2518,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432482640105925e-05,
+      "loss": 4.2397,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316456832352465e-05,
+      "loss": 4.2518,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308070884841945e-05,
+      "loss": 4.2546,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299684937331425e-05,
+      "loss": 4.2281,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42912989898209e-05,
+      "loss": 4.2398,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428291304231038e-05,
+      "loss": 4.2338,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274527094799865e-05,
+      "loss": 4.2431,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266141147289345e-05,
+      "loss": 4.2436,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257771578582554e-05,
+      "loss": 4.2369,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249385631072034e-05,
+      "loss": 4.2374,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240999683561514e-05,
+      "loss": 4.2329,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232613736050994e-05,
+      "loss": 4.2355,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224227788540474e-05,
+      "loss": 4.2333,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215841841029954e-05,
+      "loss": 4.2434,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207455893519433e-05,
+      "loss": 4.247,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199069946008913e-05,
+      "loss": 4.2359,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419068399849839e-05,
+      "loss": 4.23,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41823144297916e-05,
+      "loss": 4.2334,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417392848228108e-05,
+      "loss": 4.2349,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416554253477056e-05,
+      "loss": 4.2416,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415715658726004e-05,
+      "loss": 4.2357,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414880339735699e-05,
+      "loss": 4.2297,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414041744984647e-05,
+      "loss": 4.2239,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413203150233595e-05,
+      "loss": 4.236,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412364555482543e-05,
+      "loss": 4.22,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411525960731491e-05,
+      "loss": 4.2218,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106890038608116e-05,
+      "loss": 4.2178,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098504091097596e-05,
+      "loss": 4.2306,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4090118143587076e-05,
+      "loss": 4.221,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081732196076556e-05,
+      "loss": 4.2416,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073346248566036e-05,
+      "loss": 4.2187,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064960301055516e-05,
+      "loss": 4.2258,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056574353544996e-05,
+      "loss": 4.2277,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404818840603448e-05,
+      "loss": 4.2149,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403981883732769e-05,
+      "loss": 4.2374,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403143288981717e-05,
+      "loss": 4.2326,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402304694230665e-05,
+      "loss": 4.2283,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401466099479613e-05,
+      "loss": 4.2268,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400629142608934e-05,
+      "loss": 4.2205,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399790547857882e-05,
+      "loss": 4.2335,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39895195310683e-05,
+      "loss": 4.2146,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398113358355778e-05,
+      "loss": 4.2378,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397276401485099e-05,
+      "loss": 4.2237,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396437806734047e-05,
+      "loss": 4.2244,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395599211982995e-05,
+      "loss": 4.2185,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947606172319436e-05,
+      "loss": 4.2187,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393922022480891e-05,
+      "loss": 4.2128,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930850656102125e-05,
+      "loss": 4.2289,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922464708591605e-05,
+      "loss": 4.2155,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914078761081085e-05,
+      "loss": 4.2327,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390569281357056e-05,
+      "loss": 4.2268,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897323244863774e-05,
+      "loss": 4.2241,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888937297353254e-05,
+      "loss": 4.2118,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388055134984273e-05,
+      "loss": 4.2277,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387216540233221e-05,
+      "loss": 4.2242,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386377945482169e-05,
+      "loss": 4.2297,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38554098861149e-05,
+      "loss": 4.2188,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384702393860438e-05,
+      "loss": 4.2148,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38386543698976e-05,
+      "loss": 4.2239,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383026842238708e-05,
+      "loss": 4.2273,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382188247487656e-05,
+      "loss": 4.2101,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381349652736603e-05,
+      "loss": 4.2233,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380511057985551e-05,
+      "loss": 4.2111,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379672463234499e-05,
+      "loss": 4.2133,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378833868483447e-05,
+      "loss": 4.2239,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377995273732395e-05,
+      "loss": 4.2156,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377158316861716e-05,
+      "loss": 4.2117,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376319722110664e-05,
+      "loss": 4.2126,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375481127359613e-05,
+      "loss": 4.2163,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.205813884735107,
+      "eval_runtime": 296.6556,
+      "eval_samples_per_second": 1286.31,
+      "eval_steps_per_second": 40.198,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374642532608561e-05,
+      "loss": 4.2157,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373805575737882e-05,
+      "loss": 4.2106,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37296698098683e-05,
+      "loss": 4.2259,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372128386235778e-05,
+      "loss": 4.2167,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371289791484726e-05,
+      "loss": 4.2186,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3704528346140466e-05,
+      "loss": 4.2154,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3696142398629946e-05,
+      "loss": 4.2118,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3687756451119426e-05,
+      "loss": 4.204,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3679370503608906e-05,
+      "loss": 4.2205,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670984556098386e-05,
+      "loss": 4.2241,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662614987391595e-05,
+      "loss": 4.2163,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365424541868481e-05,
+      "loss": 4.2255,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645875849978027e-05,
+      "loss": 4.2014,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637489902467506e-05,
+      "loss": 4.2134,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362910395495698e-05,
+      "loss": 4.2196,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362071800744646e-05,
+      "loss": 4.1984,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361233205993594e-05,
+      "loss": 4.207,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360394611242542e-05,
+      "loss": 4.2044,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35955601649149e-05,
+      "loss": 4.208,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358717421740438e-05,
+      "loss": 4.2218,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357878826989386e-05,
+      "loss": 4.2109,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357040232238334e-05,
+      "loss": 4.212,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356201637487282e-05,
+      "loss": 4.1993,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35536304273623e-05,
+      "loss": 4.216,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354524447985178e-05,
+      "loss": 4.2011,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3536874911144995e-05,
+      "loss": 4.1993,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528488963634475e-05,
+      "loss": 4.2008,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520103016123955e-05,
+      "loss": 4.2063,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511717068613435e-05,
+      "loss": 4.1936,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503347499906644e-05,
+      "loss": 4.206,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494961552396124e-05,
+      "loss": 4.2063,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486575604885604e-05,
+      "loss": 4.2077,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478189657375084e-05,
+      "loss": 4.2027,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346982008866829e-05,
+      "loss": 4.2033,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346143414115777e-05,
+      "loss": 4.2107,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345304819364725e-05,
+      "loss": 4.2025,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344466224613673e-05,
+      "loss": 4.1909,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343629267742995e-05,
+      "loss": 4.1955,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342790672991943e-05,
+      "loss": 4.1891,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341952078240891e-05,
+      "loss": 4.1993,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341113483489839e-05,
+      "loss": 4.198,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34027652661916e-05,
+      "loss": 4.2042,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339437931868108e-05,
+      "loss": 4.1947,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338599337117056e-05,
+      "loss": 4.1979,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337760742366004e-05,
+      "loss": 4.2054,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336922147614951e-05,
+      "loss": 4.1957,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336083552863899e-05,
+      "loss": 4.204,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352465959932207e-05,
+      "loss": 4.1918,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344080012421687e-05,
+      "loss": 4.1835,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3335694064911167e-05,
+      "loss": 4.2124,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327308117400646e-05,
+      "loss": 4.1936,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331893854869386e-05,
+      "loss": 4.1929,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3310552601183336e-05,
+      "loss": 4.1904,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3302166653672815e-05,
+      "loss": 4.1984,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329379708496603e-05,
+      "loss": 4.1808,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328541113745551e-05,
+      "loss": 4.1993,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3277025189944984e-05,
+      "loss": 4.1928,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3268639242434464e-05,
+      "loss": 4.1907,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260253294923944e-05,
+      "loss": 4.2075,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3251867347413424e-05,
+      "loss": 4.1839,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324349777870664e-05,
+      "loss": 4.1864,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323511183119612e-05,
+      "loss": 4.1981,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32267258836856e-05,
+      "loss": 4.1792,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321833993617508e-05,
+      "loss": 4.1823,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320995398866456e-05,
+      "loss": 4.1854,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320156804115404e-05,
+      "loss": 4.1938,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319318209364352e-05,
+      "loss": 4.1814,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184796146133e-05,
+      "loss": 4.179,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317641019862248e-05,
+      "loss": 4.1836,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316804062991569e-05,
+      "loss": 4.1954,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315965468240517e-05,
+      "loss": 4.192,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315126873489465e-05,
+      "loss": 4.1932,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314288278738413e-05,
+      "loss": 4.1948,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313449683987361e-05,
+      "loss": 4.1958,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3126127271166825e-05,
+      "loss": 4.206,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3117741323656305e-05,
+      "loss": 4.1742,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109355376145785e-05,
+      "loss": 4.1962,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3100985807438994e-05,
+      "loss": 4.1976,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3092599859928474e-05,
+      "loss": 4.1698,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3084213912417954e-05,
+      "loss": 4.1973,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075827964907434e-05,
+      "loss": 4.1823,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067442017396913e-05,
+      "loss": 4.1963,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3059056069886393e-05,
+      "loss": 4.1988,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050670122375873e-05,
+      "loss": 4.1761,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042284174865347e-05,
+      "loss": 4.1815,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033898227354827e-05,
+      "loss": 4.1809,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302552865864804e-05,
+      "loss": 4.1894,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301714271113753e-05,
+      "loss": 4.1862,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008756763627e-05,
+      "loss": 4.1857,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300037081611648e-05,
+      "loss": 4.1782,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299198486860596e-05,
+      "loss": 4.1805,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298361529989917e-05,
+      "loss": 4.182,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297522935238865e-05,
+      "loss": 4.1775,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296684340487813e-05,
+      "loss": 4.1928,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295845745736761e-05,
+      "loss": 4.1892,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295008788866082e-05,
+      "loss": 4.1895,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29417019411503e-05,
+      "loss": 4.1722,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933332372443516e-05,
+      "loss": 4.179,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924946424932996e-05,
+      "loss": 4.1799,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916560477422476e-05,
+      "loss": 4.1844,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908174529911956e-05,
+      "loss": 4.1852,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2899788582401436e-05,
+      "loss": 4.178,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2891402634890916e-05,
+      "loss": 4.171,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2883016687380396e-05,
+      "loss": 4.1821,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2874630739869876e-05,
+      "loss": 4.1704,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866261171163085e-05,
+      "loss": 4.1669,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2857875223652565e-05,
+      "loss": 4.1623,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849489276142045e-05,
+      "loss": 4.1781,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2841103328631525e-05,
+      "loss": 4.174,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2832733759924734e-05,
+      "loss": 4.1862,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282434781241422e-05,
+      "loss": 4.1683,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28159618649037e-05,
+      "loss": 4.1731,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280757591739318e-05,
+      "loss": 4.1767,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279918996988266e-05,
+      "loss": 4.1632,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279080402237214e-05,
+      "loss": 4.1856,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278243445366535e-05,
+      "loss": 4.1796,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277406488495856e-05,
+      "loss": 4.1745,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276567893744804e-05,
+      "loss": 4.1752,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275729298993752e-05,
+      "loss": 4.1711,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748907042427e-05,
+      "loss": 4.1807,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274052109491648e-05,
+      "loss": 4.1594,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273213514740596e-05,
+      "loss": 4.1886,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272374919989544e-05,
+      "loss": 4.1697,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271536325238492e-05,
+      "loss": 4.1716,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27069773048744e-05,
+      "loss": 4.1704,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269862411497134e-05,
+      "loss": 4.1657,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269023816746082e-05,
+      "loss": 4.1673,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26818522199503e-05,
+      "loss": 4.1765,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267346627243978e-05,
+      "loss": 4.168,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266508032492926e-05,
+      "loss": 4.1802,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265669437741874e-05,
+      "loss": 4.1725,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264830842990822e-05,
+      "loss": 4.1777,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26399224823977e-05,
+      "loss": 4.1585,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263155291369091e-05,
+      "loss": 4.1808,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262316696618039e-05,
+      "loss": 4.1703,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261478101866987e-05,
+      "loss": 4.1784,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260639507115935e-05,
+      "loss": 4.1696,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259800912364883e-05,
+      "loss": 4.1651,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258963955494205e-05,
+      "loss": 4.1791,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258125360743153e-05,
+      "loss": 4.1778,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257286765992101e-05,
+      "loss": 4.1591,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256448171241048e-05,
+      "loss": 4.1703,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255609576489996e-05,
+      "loss": 4.1663,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254770981738944e-05,
+      "loss": 4.1633,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253932386987892e-05,
+      "loss": 4.1747,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253095430117213e-05,
+      "loss": 4.1656,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252256835366161e-05,
+      "loss": 4.1698,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251418240615109e-05,
+      "loss": 4.1565,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2505796458640576e-05,
+      "loss": 4.1714,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.164942741394043,
+      "eval_runtime": 290.7804,
+      "eval_samples_per_second": 1312.299,
+      "eval_steps_per_second": 41.01,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497410511130056e-05,
+      "loss": 4.1691,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2489024563619536e-05,
+      "loss": 4.1591,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480638616109016e-05,
+      "loss": 4.1758,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472252668598496e-05,
+      "loss": 4.1696,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2463866721087976e-05,
+      "loss": 4.1694,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455480773577456e-05,
+      "loss": 4.1661,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447094826066936e-05,
+      "loss": 4.1581,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2438708878556416e-05,
+      "loss": 4.1608,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430322931045896e-05,
+      "loss": 4.1687,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242193698353537e-05,
+      "loss": 4.1775,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241355103602485e-05,
+      "loss": 4.1672,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240516508851433e-05,
+      "loss": 4.1763,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2396795519807545e-05,
+      "loss": 4.1583,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388409572297025e-05,
+      "loss": 4.1595,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380023624786505e-05,
+      "loss": 4.1719,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371637677275985e-05,
+      "loss": 4.1529,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363251729765465e-05,
+      "loss": 4.1589,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354865782254945e-05,
+      "loss": 4.1552,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346479834744425e-05,
+      "loss": 4.1624,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2338093887233905e-05,
+      "loss": 4.1689,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329707939723385e-05,
+      "loss": 4.1684,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321321992212865e-05,
+      "loss": 4.165,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312936044702345e-05,
+      "loss": 4.1495,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304550097191825e-05,
+      "loss": 4.1688,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296180528485034e-05,
+      "loss": 4.1566,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228781095977824e-05,
+      "loss": 4.1533,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227942501226772e-05,
+      "loss": 4.1485,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227103906475721e-05,
+      "loss": 4.1556,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226265311724669e-05,
+      "loss": 4.1561,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225426716973617e-05,
+      "loss": 4.1521,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224588122222565e-05,
+      "loss": 4.1612,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223751165351886e-05,
+      "loss": 4.1604,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222912570600834e-05,
+      "loss": 4.1565,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222073975849782e-05,
+      "loss": 4.1618,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22123538109873e-05,
+      "loss": 4.1591,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220396786347678e-05,
+      "loss": 4.1562,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219558191596626e-05,
+      "loss": 4.1521,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218719596845573e-05,
+      "loss": 4.1426,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217881002094521e-05,
+      "loss": 4.1417,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217044045223843e-05,
+      "loss": 4.1527,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216205450472791e-05,
+      "loss": 4.1532,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215366855721739e-05,
+      "loss": 4.1553,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214528260970687e-05,
+      "loss": 4.1498,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213689666219635e-05,
+      "loss": 4.15,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128527093489556e-05,
+      "loss": 4.1621,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120141145979036e-05,
+      "loss": 4.1497,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111755198468516e-05,
+      "loss": 4.1595,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103369250957996e-05,
+      "loss": 4.1455,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094999682251205e-05,
+      "loss": 4.1347,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2086613734740685e-05,
+      "loss": 4.1669,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078227787230165e-05,
+      "loss": 4.1522,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206985821852338e-05,
+      "loss": 4.1445,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206147227101286e-05,
+      "loss": 4.1473,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205308632350234e-05,
+      "loss": 4.149,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204470037599182e-05,
+      "loss": 4.1404,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20363144284813e-05,
+      "loss": 4.1484,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202792848097078e-05,
+      "loss": 4.1477,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201954253346026e-05,
+      "loss": 4.1485,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201115658594974e-05,
+      "loss": 4.1607,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200278701724295e-05,
+      "loss": 4.1382,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199440106973243e-05,
+      "loss": 4.1415,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198601512222191e-05,
+      "loss": 4.1564,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197762917471139e-05,
+      "loss": 4.1342,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19692596060046e-05,
+      "loss": 4.1344,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1960873658494085e-05,
+      "loss": 4.1421,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1952504089787294e-05,
+      "loss": 4.1475,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1944118142276774e-05,
+      "loss": 4.135,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1935732194766254e-05,
+      "loss": 4.1379,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927346247255734e-05,
+      "loss": 4.1359,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918960299745214e-05,
+      "loss": 4.1571,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910574352234694e-05,
+      "loss": 4.1484,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902188404724174e-05,
+      "loss": 4.1489,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1893802457213654e-05,
+      "loss": 4.1491,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885416509703134e-05,
+      "loss": 4.1515,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187704694099634e-05,
+      "loss": 4.1629,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186866099348582e-05,
+      "loss": 4.1329,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18602750459753e-05,
+      "loss": 4.1493,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185188909846478e-05,
+      "loss": 4.1548,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184350315095426e-05,
+      "loss": 4.1296,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183514996105121e-05,
+      "loss": 4.1491,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182676401354069e-05,
+      "loss": 4.1415,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181837806603017e-05,
+      "loss": 4.1526,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180999211851965e-05,
+      "loss": 4.1558,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180160617100913e-05,
+      "loss": 4.1376,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179322022349861e-05,
+      "loss": 4.1336,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178483427598809e-05,
+      "loss": 4.1366,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177644832847757e-05,
+      "loss": 4.1472,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176806238096704e-05,
+      "loss": 4.1412,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175970919106399e-05,
+      "loss": 4.1429,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175132324355347e-05,
+      "loss": 4.1397,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174293729604295e-05,
+      "loss": 4.1352,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173455134853243e-05,
+      "loss": 4.1421,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172616540102191e-05,
+      "loss": 4.1358,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171777945351139e-05,
+      "loss": 4.1487,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709393506000866e-05,
+      "loss": 4.145,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701007558490346e-05,
+      "loss": 4.1467,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692621610979826e-05,
+      "loss": 4.1313,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168425204227304e-05,
+      "loss": 4.1371,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1675866094762515e-05,
+      "loss": 4.1417,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667480147251995e-05,
+      "loss": 4.1424,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1659094199741474e-05,
+      "loss": 4.1421,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165072463103469e-05,
+      "loss": 4.131,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164233868352417e-05,
+      "loss": 4.1349,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163395273601365e-05,
+      "loss": 4.1372,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162556678850313e-05,
+      "loss": 4.1282,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161718084099261e-05,
+      "loss": 4.1342,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160881127228582e-05,
+      "loss": 4.1145,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1600441703579035e-05,
+      "loss": 4.1377,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159205575606851e-05,
+      "loss": 4.1339,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158366980855799e-05,
+      "loss": 4.1396,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157528386104747e-05,
+      "loss": 4.1284,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156689791353695e-05,
+      "loss": 4.1333,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155851196602643e-05,
+      "loss": 4.137,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155012601851591e-05,
+      "loss": 4.1188,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1541740071005395e-05,
+      "loss": 4.1477,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533370502298604e-05,
+      "loss": 4.1365,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152500093359181e-05,
+      "loss": 4.1347,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151661498608129e-05,
+      "loss": 4.1379,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150822903857077e-05,
+      "loss": 4.1218,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149984309106025e-05,
+      "loss": 4.1434,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149145714354973e-05,
+      "loss": 4.1232,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148307119603921e-05,
+      "loss": 4.1458,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147468524852869e-05,
+      "loss": 4.1253,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146629930101817e-05,
+      "loss": 4.1333,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145791335350765e-05,
+      "loss": 4.1276,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144952740599713e-05,
+      "loss": 4.1273,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144114145848661e-05,
+      "loss": 4.1281,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143277188977983e-05,
+      "loss": 4.1309,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142438594226931e-05,
+      "loss": 4.1309,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141599999475879e-05,
+      "loss": 4.1389,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140761404724827e-05,
+      "loss": 4.1305,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139922809973775e-05,
+      "loss": 4.1399,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139084215222723e-05,
+      "loss": 4.1194,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13824562047167e-05,
+      "loss": 4.14,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137407025720618e-05,
+      "loss": 4.1348,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136568430969566e-05,
+      "loss": 4.1346,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135731474098887e-05,
+      "loss": 4.1296,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134892879347835e-05,
+      "loss": 4.1306,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134054284596783e-05,
+      "loss": 4.1328,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133215689845732e-05,
+      "loss": 4.1403,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13237709509468e-05,
+      "loss": 4.118,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131538500343628e-05,
+      "loss": 4.1332,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307015434729486e-05,
+      "loss": 4.1258,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298629487218966e-05,
+      "loss": 4.1257,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290243539708446e-05,
+      "loss": 4.1311,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281857592197926e-05,
+      "loss": 4.128,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273488023491135e-05,
+      "loss": 4.1283,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1265102075980615e-05,
+      "loss": 4.1178,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1256716128470095e-05,
+      "loss": 4.1359,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.133894920349121,
+      "eval_runtime": 315.3098,
+      "eval_samples_per_second": 1210.21,
+      "eval_steps_per_second": 37.82,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1248330180959575e-05,
+      "loss": 4.1254,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1239944233449055e-05,
+      "loss": 4.1212,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123157466474227e-05,
+      "loss": 4.1343,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122320509603548e-05,
+      "loss": 4.1298,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121481914852496e-05,
+      "loss": 4.1292,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120643320101444e-05,
+      "loss": 4.1306,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119804725350392e-05,
+      "loss": 4.1213,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11896613059934e-05,
+      "loss": 4.1196,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118127535848288e-05,
+      "loss": 4.1333,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117290578977609e-05,
+      "loss": 4.1372,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116451984226557e-05,
+      "loss": 4.1298,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115613389475505e-05,
+      "loss": 4.1363,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114776432604826e-05,
+      "loss": 4.1234,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113937837853774e-05,
+      "loss": 4.1221,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113099243102722e-05,
+      "loss": 4.1299,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122606483516704e-05,
+      "loss": 4.1122,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114220536006184e-05,
+      "loss": 4.1222,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105834588495664e-05,
+      "loss": 4.1159,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1097448640985144e-05,
+      "loss": 4.1262,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1089062693474624e-05,
+      "loss": 4.129,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1080676745964104e-05,
+      "loss": 4.1343,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072290798453584e-05,
+      "loss": 4.1261,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063904850943064e-05,
+      "loss": 4.111,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105551890343254e-05,
+      "loss": 4.1306,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104713295592202e-05,
+      "loss": 4.1165,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103876338721523e-05,
+      "loss": 4.1199,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030377439704706e-05,
+      "loss": 4.1108,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021991492194186e-05,
+      "loss": 4.1211,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101360554468367e-05,
+      "loss": 4.1149,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100521959717315e-05,
+      "loss": 4.1204,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099685002846636e-05,
+      "loss": 4.1202,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098846408095584e-05,
+      "loss": 4.1286,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098007813344532e-05,
+      "loss": 4.1212,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09716921859348e-05,
+      "loss": 4.1185,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096332261722801e-05,
+      "loss": 4.1224,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095493666971749e-05,
+      "loss": 4.1196,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094656710101071e-05,
+      "loss": 4.1189,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093818115350018e-05,
+      "loss": 4.1027,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092979520598966e-05,
+      "loss": 4.1083,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092140925847914e-05,
+      "loss": 4.1131,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0913023310968627e-05,
+      "loss": 4.122,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0904637363458107e-05,
+      "loss": 4.1195,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896251415947586e-05,
+      "loss": 4.113,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0887865468437066e-05,
+      "loss": 4.113,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879495899730276e-05,
+      "loss": 4.1252,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0871109952219755e-05,
+      "loss": 4.1113,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0862724004709235e-05,
+      "loss": 4.1309,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854338057198715e-05,
+      "loss": 4.1084,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0845952109688195e-05,
+      "loss": 4.0942,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837582540981404e-05,
+      "loss": 4.1352,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829196593470884e-05,
+      "loss": 4.1141,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0820810645960364e-05,
+      "loss": 4.1136,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0812424698449844e-05,
+      "loss": 4.1113,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080405512974306e-05,
+      "loss": 4.1105,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079566918223254e-05,
+      "loss": 4.1071,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078728323472202e-05,
+      "loss": 4.1117,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077891366601523e-05,
+      "loss": 4.1128,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077052771850471e-05,
+      "loss": 4.1138,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076214177099419e-05,
+      "loss": 4.1254,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075375582348367e-05,
+      "loss": 4.102,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074536987597315e-05,
+      "loss": 4.1029,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073698392846263e-05,
+      "loss": 4.1234,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072859798095211e-05,
+      "loss": 4.0972,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072021203344159e-05,
+      "loss": 4.0974,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07118424647348e-05,
+      "loss": 4.1076,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0703472896028014e-05,
+      "loss": 4.1138,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0695086948517494e-05,
+      "loss": 4.0976,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0686701001006974e-05,
+      "loss": 4.1067,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678315053496454e-05,
+      "loss": 4.0994,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0669929105985934e-05,
+      "loss": 4.1189,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0661543158475414e-05,
+      "loss": 4.1159,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653157210964894e-05,
+      "loss": 4.1091,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06447876422581e-05,
+      "loss": 4.1154,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063640169474758e-05,
+      "loss": 4.1175,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062801574723706e-05,
+      "loss": 4.1295,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061962979972654e-05,
+      "loss": 4.096,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611243852216016e-05,
+      "loss": 4.1171,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06028579047055e-05,
+      "loss": 4.1137,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059447195719498e-05,
+      "loss": 4.0985,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05861023884882e-05,
+      "loss": 4.1117,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057771644097767e-05,
+      "loss": 4.1109,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056933049346715e-05,
+      "loss": 4.1168,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056094454595663e-05,
+      "loss": 4.1219,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055255859844611e-05,
+      "loss": 4.1011,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054417265093559e-05,
+      "loss": 4.0975,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053578670342507e-05,
+      "loss": 4.1056,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052740075591455e-05,
+      "loss": 4.1126,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051903118720776e-05,
+      "loss": 4.1114,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051066161850097e-05,
+      "loss": 4.1084,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050227567099045e-05,
+      "loss": 4.1011,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0493889723479936e-05,
+      "loss": 4.1035,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0485503775969416e-05,
+      "loss": 4.1072,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477117828458896e-05,
+      "loss": 4.1053,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0468731880948376e-05,
+      "loss": 4.1146,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460345933437856e-05,
+      "loss": 4.1142,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0451959985927336e-05,
+      "loss": 4.1092,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443590417220545e-05,
+      "loss": 4.0982,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435204469710025e-05,
+      "loss": 4.1065,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0426818522199505e-05,
+      "loss": 4.1005,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418432574688985e-05,
+      "loss": 4.1088,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0410046627178465e-05,
+      "loss": 4.1112,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0401677058471674e-05,
+      "loss": 4.1016,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0393291110961154e-05,
+      "loss": 4.0972,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0384905163450634e-05,
+      "loss": 4.0983,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037651921594012e-05,
+      "loss": 4.1,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03681332684296e-05,
+      "loss": 4.0983,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035976369972281e-05,
+      "loss": 4.0839,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035137775221229e-05,
+      "loss": 4.1039,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034299180470177e-05,
+      "loss": 4.1035,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033460585719125e-05,
+      "loss": 4.1023,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032623628848446e-05,
+      "loss": 4.0917,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031785034097394e-05,
+      "loss": 4.1073,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030946439346342e-05,
+      "loss": 4.0991,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03010784459529e-05,
+      "loss": 4.087,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029270887724611e-05,
+      "loss": 4.1108,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028432292973559e-05,
+      "loss": 4.1016,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0275936982225074e-05,
+      "loss": 4.1012,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267551034714554e-05,
+      "loss": 4.1082,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025916508720403e-05,
+      "loss": 4.0891,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025079551849724e-05,
+      "loss": 4.1096,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024242594979045e-05,
+      "loss": 4.0943,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023404000227993e-05,
+      "loss": 4.1074,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022565405476941e-05,
+      "loss": 4.0933,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021726810725889e-05,
+      "loss": 4.1011,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020888215974837e-05,
+      "loss": 4.0948,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020049621223785e-05,
+      "loss": 4.0933,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0192110264727325e-05,
+      "loss": 4.0895,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018372431721681e-05,
+      "loss": 4.0994,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017533836970629e-05,
+      "loss": 4.0995,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01669688009995e-05,
+      "loss": 4.1073,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015858285348898e-05,
+      "loss": 4.094,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015019690597846e-05,
+      "loss": 4.1095,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014181095846794e-05,
+      "loss": 4.0847,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013344138976115e-05,
+      "loss": 4.1085,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012505544225063e-05,
+      "loss": 4.1019,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011666949474011e-05,
+      "loss": 4.1029,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010828354722959e-05,
+      "loss": 4.0968,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00999139785228e-05,
+      "loss": 4.0986,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009152803101228e-05,
+      "loss": 4.1011,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0083142083501766e-05,
+      "loss": 4.1108,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0074756135991246e-05,
+      "loss": 4.0864,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0066386567284455e-05,
+      "loss": 4.1001,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0058000619773935e-05,
+      "loss": 4.0919,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0049614672263415e-05,
+      "loss": 4.0992,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041228724752895e-05,
+      "loss": 4.0931,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0032859156046104e-05,
+      "loss": 4.0944,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024473208535584e-05,
+      "loss": 4.1004,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016087261025064e-05,
+      "loss": 4.0845,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0007701313514544e-05,
+      "loss": 4.1036,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.10920524597168,
+      "eval_runtime": 315.3389,
+      "eval_samples_per_second": 1210.098,
+      "eval_steps_per_second": 37.816,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999315366004023e-05,
+      "loss": 4.0948,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9990929418493503e-05,
+      "loss": 4.0915,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9982543470982983e-05,
+      "loss": 4.0996,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997415752347246e-05,
+      "loss": 4.0976,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996577157596194e-05,
+      "loss": 4.1005,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995738562845143e-05,
+      "loss": 4.0933,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994899968094091e-05,
+      "loss": 4.0911,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994061373343039e-05,
+      "loss": 4.0925,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993222778591986e-05,
+      "loss": 4.0965,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992384183840934e-05,
+      "loss": 4.1037,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991545589089882e-05,
+      "loss": 4.1004,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99070699433883e-05,
+      "loss": 4.1037,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989868399587778e-05,
+      "loss": 4.0881,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989029804836726e-05,
+      "loss": 4.0924,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988191210085674e-05,
+      "loss": 4.0992,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987352615334622e-05,
+      "loss": 4.0827,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98651402058357e-05,
+      "loss": 4.0901,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985675425832518e-05,
+      "loss": 4.0847,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984836831081466e-05,
+      "loss": 4.0913,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983999874210788e-05,
+      "loss": 4.096,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983161279459736e-05,
+      "loss": 4.1049,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982322684708684e-05,
+      "loss": 4.0934,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981484089957632e-05,
+      "loss": 4.0818,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98064549520658e-05,
+      "loss": 4.0947,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979806900455528e-05,
+      "loss": 4.0892,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978968305704475e-05,
+      "loss": 4.0876,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978129710953423e-05,
+      "loss": 4.0786,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977292754082745e-05,
+      "loss": 4.0901,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976454159331693e-05,
+      "loss": 4.0874,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97561556458064e-05,
+      "loss": 4.083,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9747786077099616e-05,
+      "loss": 4.0873,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973941650839283e-05,
+      "loss": 4.0965,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973103056088231e-05,
+      "loss": 4.0879,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972264461337179e-05,
+      "loss": 4.0889,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971425866586127e-05,
+      "loss": 4.0913,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970587271835075e-05,
+      "loss": 4.0925,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697486770840225e-05,
+      "loss": 4.084,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689100823329705e-05,
+      "loss": 4.0725,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9680714875819185e-05,
+      "loss": 4.0787,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672328928308665e-05,
+      "loss": 4.0824,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9663942980798145e-05,
+      "loss": 4.0866,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9655557033287625e-05,
+      "loss": 4.0876,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9647171085777105e-05,
+      "loss": 4.0841,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9638785138266585e-05,
+      "loss": 4.0848,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96304155695598e-05,
+      "loss": 4.0921,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962202962204928e-05,
+      "loss": 4.0807,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961364367453876e-05,
+      "loss": 4.0971,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960525772702824e-05,
+      "loss": 4.0819,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959688815832145e-05,
+      "loss": 4.0618,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958850221081093e-05,
+      "loss": 4.1048,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958013264210414e-05,
+      "loss": 4.0793,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957174669459362e-05,
+      "loss": 4.087,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95633607470831e-05,
+      "loss": 4.0796,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955497479957258e-05,
+      "loss": 4.0773,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954658885206206e-05,
+      "loss": 4.0777,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953820290455154e-05,
+      "loss": 4.0779,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952981695704102e-05,
+      "loss": 4.0853,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9521447388334234e-05,
+      "loss": 4.083,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9513061440823714e-05,
+      "loss": 4.0958,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9504675493313194e-05,
+      "loss": 4.0727,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496289545802674e-05,
+      "loss": 4.0726,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9487903598292154e-05,
+      "loss": 4.0948,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479517650781634e-05,
+      "loss": 4.0656,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471131703271114e-05,
+      "loss": 4.0671,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946274575576059e-05,
+      "loss": 4.0778,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94543761870538e-05,
+      "loss": 4.0822,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944599023954328e-05,
+      "loss": 4.0699,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943762067083649e-05,
+      "loss": 4.0768,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942925110212971e-05,
+      "loss": 4.0692,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942086515461919e-05,
+      "loss": 4.0884,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941247920710867e-05,
+      "loss": 4.0855,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940409325959815e-05,
+      "loss": 4.0812,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939570731208763e-05,
+      "loss": 4.0856,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938732136457711e-05,
+      "loss": 4.0883,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937893541706659e-05,
+      "loss": 4.0977,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937054946955606e-05,
+      "loss": 4.0684,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936216352204554e-05,
+      "loss": 4.0887,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935377757453502e-05,
+      "loss": 4.0788,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93453916270245e-05,
+      "loss": 4.0721,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933700567951398e-05,
+      "loss": 4.0826,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932863611080719e-05,
+      "loss": 4.083,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932025016329668e-05,
+      "loss": 4.0848,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931186421578616e-05,
+      "loss": 4.0952,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303478268275637e-05,
+      "loss": 4.0701,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295092320765117e-05,
+      "loss": 4.0657,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9286706373254597e-05,
+      "loss": 4.0778,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278320425744076e-05,
+      "loss": 4.0847,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9269934478233556e-05,
+      "loss": 4.0825,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9261581288330495e-05,
+      "loss": 4.0796,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9253195340819975e-05,
+      "loss": 4.0701,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9244809393309455e-05,
+      "loss": 4.0745,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236423445798935e-05,
+      "loss": 4.0811,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9228037498288414e-05,
+      "loss": 4.0727,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9219651550777894e-05,
+      "loss": 4.0842,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211265603267374e-05,
+      "loss": 4.0856,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920287965575686e-05,
+      "loss": 4.0796,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919451008705007e-05,
+      "loss": 4.0718,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918612413953955e-05,
+      "loss": 4.074,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917775457083276e-05,
+      "loss": 4.074,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916936862332224e-05,
+      "loss": 4.0777,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916098267581172e-05,
+      "loss": 4.0816,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91525967283012e-05,
+      "loss": 4.0754,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914421078079068e-05,
+      "loss": 4.0722,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913582483328016e-05,
+      "loss": 4.0626,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912743888576964e-05,
+      "loss": 4.0767,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911905293825912e-05,
+      "loss": 4.0678,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911068336955233e-05,
+      "loss": 4.0542,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102297422041815e-05,
+      "loss": 4.0799,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9093911474531295e-05,
+      "loss": 4.0746,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9085525527020775e-05,
+      "loss": 4.0754,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077155958313984e-05,
+      "loss": 4.0628,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068770010803464e-05,
+      "loss": 4.0769,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906040044209667e-05,
+      "loss": 4.0721,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905201449458615e-05,
+      "loss": 4.0612,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904362854707563e-05,
+      "loss": 4.0856,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903524259956511e-05,
+      "loss": 4.0753,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902685665205459e-05,
+      "loss": 4.0733,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90184870833478e-05,
+      "loss": 4.0788,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901010113583728e-05,
+      "loss": 4.0591,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900171518832676e-05,
+      "loss": 4.0793,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899334561961998e-05,
+      "loss": 4.0694,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898495967210946e-05,
+      "loss": 4.076,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897657372459894e-05,
+      "loss": 4.0674,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896818777708842e-05,
+      "loss": 4.0773,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89598018295779e-05,
+      "loss": 4.0698,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895141588206737e-05,
+      "loss": 4.0645,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894302993455685e-05,
+      "loss": 4.0625,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8934660365850066e-05,
+      "loss": 4.0748,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892627441833954e-05,
+      "loss": 4.0671,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891788847082902e-05,
+      "loss": 4.0787,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8909502523318506e-05,
+      "loss": 4.0679,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8901132954611715e-05,
+      "loss": 4.0844,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892747007101195e-05,
+      "loss": 4.0585,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884361059590675e-05,
+      "loss": 4.0796,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8875975112080155e-05,
+      "loss": 4.0759,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8867589164569635e-05,
+      "loss": 4.0768,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8859203217059115e-05,
+      "loss": 4.0685,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8850817269548595e-05,
+      "loss": 4.0706,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8842431322038075e-05,
+      "loss": 4.0727,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8834061753331284e-05,
+      "loss": 4.0899,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8825675805820764e-05,
+      "loss": 4.057,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817289858310244e-05,
+      "loss": 4.0697,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8808903910799724e-05,
+      "loss": 4.065,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8800517963289204e-05,
+      "loss": 4.0753,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879214839458242e-05,
+      "loss": 4.0635,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87837624470719e-05,
+      "loss": 4.0684,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877537649956138e-05,
+      "loss": 4.0704,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876700693085459e-05,
+      "loss": 4.0641,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875862098334407e-05,
+      "loss": 4.0733,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.089617729187012,
+      "eval_runtime": 319.4369,
+      "eval_samples_per_second": 1194.574,
+      "eval_steps_per_second": 37.331,
+      "step": 686880
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.849713077366573e+17,
+  "trial_name": null,
+  "trial_params": null
+}