diff --git "a/checkpoint-839520/trainer_state.json" "b/checkpoint-839520/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-839520/trainer_state.json"
@@ -0,0 +1,9947 @@
+{
+  "best_metric": 4.075167179107666,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/lstm/2/checkpoints/checkpoint-839520",
+  "epoch": 2.025000606015738,
+  "eval_steps": 10,
+  "global_step": 839520,
+  "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": 10.8202,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5484,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0528,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9887,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9423,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9024,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.732,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.626,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5233,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4512,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3812,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.324,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2579,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1877,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1303,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0634,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0219,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9809,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9399,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9034,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8747,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8349,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8113,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7679,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.7551,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979036769104073e-05,
+      "loss": 5.7238,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.978198174353021e-05,
+      "loss": 5.7047,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.6762,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6445,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6281,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6031,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 5.5931,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5668,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.556,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5444,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5233,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5128,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4687,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4655,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.444,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.431,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4256,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4178,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.3898,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3895,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3845,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3711,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3656,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3495,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3285,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.33,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3116,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3131,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.282,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 5.2943,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.2674,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2857,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2572,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2442,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 5.2398,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.2354,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2208,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2225,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 5.196,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 5.1905,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.2036,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.1956,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1782,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1613,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1594,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 5.1609,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 5.1542,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 5.1498,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 5.1459,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 5.1466,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 5.1378,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1114,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1079,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.105,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0932,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.102,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0841,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.1024,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0772,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.062,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0656,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0663,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0635,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0498,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0448,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0374,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0354,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0255,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0204,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0195,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0248,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 5.0054,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9993,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9923,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.9881,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9944,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9798,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9731,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.9661,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9697,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9401,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9568,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9442,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9505,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9377,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9447,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9302,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9255,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9329,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9102,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.927,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9161,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9135,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.9007,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.9015,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9044,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8946,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8946,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8845,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.886,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8675,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8699,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8532,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8594,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8463,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8679,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.8729,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8578,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8355,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8457,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8458,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.8391,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.835,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8343,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8307,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8201,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8209,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.821,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8095,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8089,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.8163,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.8046,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.8039,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.8146,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7913,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.780590534210205,
+      "eval_runtime": 282.4544,
+      "eval_samples_per_second": 1350.982,
+      "eval_steps_per_second": 42.219,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.7797,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.7866,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.8002,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.7879,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7927,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.7681,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7842,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7566,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7726,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.759,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7738,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7684,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.7567,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.7508,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.739,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.7393,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.7491,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.7449,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.7371,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.755,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.7461,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.7382,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7267,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.7464,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7218,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.7312,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7283,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7209,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7103,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8499455896140044e-05,
+      "loss": 4.7113,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.7122,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.7097,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.71,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.7156,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.705,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.704,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6835,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6906,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6804,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6773,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.6885,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.6889,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6728,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.68,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.6779,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.6805,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6829,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6733,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6626,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.6686,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6564,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6658,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6535,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6639,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.6424,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.6662,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.6492,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264731259864135e-05,
+      "loss": 4.6446,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.6516,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6417,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.6336,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.6505,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822281790111527e-05,
+      "loss": 4.6192,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.626,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6375,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.6374,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.6271,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.6189,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.618,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6244,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.6292,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814737713112805e-05,
+      "loss": 4.6237,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.6251,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.6295,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.6262,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811383334108597e-05,
+      "loss": 4.6079,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810544739357545e-05,
+      "loss": 4.6072,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.6145,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.6,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.6113,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.5977,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.6223,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.6043,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.5888,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.598,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.5993,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.6047,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.5938,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.5937,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.5965,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.585,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5846,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5832,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.5913,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.5881,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5855,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.5725,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5729,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5776,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.5758,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5728,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.5676,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5619,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.5731,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870739136103276e-05,
+      "loss": 4.541,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862353188592756e-05,
+      "loss": 4.5658,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.5568,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.5639,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.5512,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828809398550675e-05,
+      "loss": 4.564,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.5522,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.55,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803667934822844e-05,
+      "loss": 4.561,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795281987312324e-05,
+      "loss": 4.5442,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.5592,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.5559,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777014052358449e-05,
+      "loss": 4.5514,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776175457607397e-05,
+      "loss": 4.5425,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775336862856345e-05,
+      "loss": 4.5456,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.551,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.5362,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.5545,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.5361,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5416,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.5353,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.5333,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5149,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767794423737997e-05,
+      "loss": 4.5284,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.5167,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.5398,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5509,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.5334,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.5174,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762762855231685e-05,
+      "loss": 4.5294,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761924260480633e-05,
+      "loss": 4.5235,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.5267,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760248708858902e-05,
+      "loss": 4.5227,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759411751988223e-05,
+      "loss": 4.522,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758573157237171e-05,
+      "loss": 4.5181,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757734562486119e-05,
+      "loss": 4.5183,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756895967735067e-05,
+      "loss": 4.5179,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756057372984015e-05,
+      "loss": 4.5224,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.5142,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754381821362284e-05,
+      "loss": 4.5073,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753543226611232e-05,
+      "loss": 4.522,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75270463186018e-05,
+      "loss": 4.5149,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751866037109128e-05,
+      "loss": 4.5058,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751027442358077e-05,
+      "loss": 4.5237,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750188847607025e-05,
+      "loss": 4.504,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.497745037078857,
+      "eval_runtime": 283.3858,
+      "eval_samples_per_second": 1346.542,
+      "eval_steps_per_second": 42.08,
+      "step": 152640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.749350252855973e-05,
+      "loss": 4.4937,
+      "step": 153088
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.748511658104921e-05,
+      "loss": 4.4989,
+      "step": 153600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7476747012342416e-05,
+      "loss": 4.5153,
+      "step": 154112
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7468361064831896e-05,
+      "loss": 4.5065,
+      "step": 154624
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7459975117321376e-05,
+      "loss": 4.5109,
+      "step": 155136
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.745158916981085e-05,
+      "loss": 4.4893,
+      "step": 155648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7443219601104065e-05,
+      "loss": 4.5082,
+      "step": 156160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7434850032397274e-05,
+      "loss": 4.4799,
+      "step": 156672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7426464084886754e-05,
+      "loss": 4.5036,
+      "step": 157184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7418078137376234e-05,
+      "loss": 4.4864,
+      "step": 157696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.740969218986572e-05,
+      "loss": 4.5039,
+      "step": 158208
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.74013062423552e-05,
+      "loss": 4.5015,
+      "step": 158720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.739295305245214e-05,
+      "loss": 4.4891,
+      "step": 159232
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.738456710494162e-05,
+      "loss": 4.4856,
+      "step": 159744
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.73761811574311e-05,
+      "loss": 4.4793,
+      "step": 160256
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.736779520992058e-05,
+      "loss": 4.4794,
+      "step": 160768
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.735940926241006e-05,
+      "loss": 4.4848,
+      "step": 161280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.735102331489954e-05,
+      "loss": 4.486,
+      "step": 161792
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.734263736738902e-05,
+      "loss": 4.4857,
+      "step": 162304
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.73342514198785e-05,
+      "loss": 4.4974,
+      "step": 162816
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.732586547236797e-05,
+      "loss": 4.4937,
+      "step": 163328
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.731747952485746e-05,
+      "loss": 4.4879,
+      "step": 163840
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.730909357734694e-05,
+      "loss": 4.4778,
+      "step": 164352
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.730070762983642e-05,
+      "loss": 4.4909,
+      "step": 164864
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.729233806112963e-05,
+      "loss": 4.4761,
+      "step": 165376
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7283968492422844e-05,
+      "loss": 4.4833,
+      "step": 165888
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7275582544912323e-05,
+      "loss": 4.4812,
+      "step": 166400
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.72671965974018e-05,
+      "loss": 4.4772,
+      "step": 166912
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7258810649891277e-05,
+      "loss": 4.4678,
+      "step": 167424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7250424702380757e-05,
+      "loss": 4.4693,
+      "step": 167936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7242038754870237e-05,
+      "loss": 4.4715,
+      "step": 168448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7233669186163446e-05,
+      "loss": 4.4747,
+      "step": 168960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7225283238652926e-05,
+      "loss": 4.4771,
+      "step": 169472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.721689729114241e-05,
+      "loss": 4.4728,
+      "step": 169984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.720852772243563e-05,
+      "loss": 4.4694,
+      "step": 170496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.72001417749251e-05,
+      "loss": 4.4699,
+      "step": 171008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.719175582741458e-05,
+      "loss": 4.4514,
+      "step": 171520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.718336987990406e-05,
+      "loss": 4.4627,
+      "step": 172032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.717498393239354e-05,
+      "loss": 4.4524,
+      "step": 172544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.716661436368675e-05,
+      "loss": 4.4528,
+      "step": 173056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.715822841617623e-05,
+      "loss": 4.4585,
+      "step": 173568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.714984246866571e-05,
+      "loss": 4.466,
+      "step": 174080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.714145652115519e-05,
+      "loss": 4.4517,
+      "step": 174592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.713307057364467e-05,
+      "loss": 4.4543,
+      "step": 175104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.712468462613415e-05,
+      "loss": 4.4626,
+      "step": 175616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.711629867862363e-05,
+      "loss": 4.454,
+      "step": 176128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7107929109916846e-05,
+      "loss": 4.4667,
+      "step": 176640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7099543162406326e-05,
+      "loss": 4.4512,
+      "step": 177152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7091157214895806e-05,
+      "loss": 4.4427,
+      "step": 177664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7082771267385286e-05,
+      "loss": 4.454,
+      "step": 178176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7074385319874766e-05,
+      "loss": 4.4441,
+      "step": 178688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7065999372364246e-05,
+      "loss": 4.4531,
+      "step": 179200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7057613424853726e-05,
+      "loss": 4.4413,
+      "step": 179712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7049227477343206e-05,
+      "loss": 4.4471,
+      "step": 180224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7040857908636415e-05,
+      "loss": 4.436,
+      "step": 180736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7032471961125895e-05,
+      "loss": 4.4506,
+      "step": 181248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7024086013615375e-05,
+      "loss": 4.4397,
+      "step": 181760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7015700066104855e-05,
+      "loss": 4.4392,
+      "step": 182272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7007330497398064e-05,
+      "loss": 4.4455,
+      "step": 182784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6998944549887544e-05,
+      "loss": 4.4334,
+      "step": 183296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.699055860237703e-05,
+      "loss": 4.4284,
+      "step": 183808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.698217265486651e-05,
+      "loss": 4.4508,
+      "step": 184320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.697380308615972e-05,
+      "loss": 4.4174,
+      "step": 184832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.69654171386492e-05,
+      "loss": 4.4214,
+      "step": 185344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.695703119113868e-05,
+      "loss": 4.4303,
+      "step": 185856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.694864524362816e-05,
+      "loss": 4.4419,
+      "step": 186368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.694025929611763e-05,
+      "loss": 4.4278,
+      "step": 186880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.693187334860711e-05,
+      "loss": 4.4216,
+      "step": 187392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.692348740109659e-05,
+      "loss": 4.4224,
+      "step": 187904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.691511783238981e-05,
+      "loss": 4.4258,
+      "step": 188416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.690673188487928e-05,
+      "loss": 4.4339,
+      "step": 188928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.689834593736877e-05,
+      "loss": 4.4302,
+      "step": 189440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.688995998985825e-05,
+      "loss": 4.429,
+      "step": 189952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.688157404234773e-05,
+      "loss": 4.4337,
+      "step": 190464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.687320447364094e-05,
+      "loss": 4.4356,
+      "step": 190976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.686481852613042e-05,
+      "loss": 4.4165,
+      "step": 191488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.68564325786199e-05,
+      "loss": 4.4163,
+      "step": 192000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.684804663110938e-05,
+      "loss": 4.4266,
+      "step": 192512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.683966068359886e-05,
+      "loss": 4.4093,
+      "step": 193024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.683127473608834e-05,
+      "loss": 4.4248,
+      "step": 193536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6822905167381546e-05,
+      "loss": 4.4113,
+      "step": 194048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6814519219871026e-05,
+      "loss": 4.432,
+      "step": 194560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6806133272360506e-05,
+      "loss": 4.4198,
+      "step": 195072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6797747324849986e-05,
+      "loss": 4.4058,
+      "step": 195584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6789361377339466e-05,
+      "loss": 4.4094,
+      "step": 196096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.678097542982895e-05,
+      "loss": 4.4168,
+      "step": 196608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.677258948231843e-05,
+      "loss": 4.4254,
+      "step": 197120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.676420353480791e-05,
+      "loss": 4.4111,
+      "step": 197632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.675583396610112e-05,
+      "loss": 4.4151,
+      "step": 198144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.67474480185906e-05,
+      "loss": 4.4066,
+      "step": 198656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.673906207108008e-05,
+      "loss": 4.4155,
+      "step": 199168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.673067612356956e-05,
+      "loss": 4.4066,
+      "step": 199680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.672229017605904e-05,
+      "loss": 4.403,
+      "step": 200192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.671390422854852e-05,
+      "loss": 4.4145,
+      "step": 200704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.670553465984173e-05,
+      "loss": 4.4124,
+      "step": 201216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.669714871233121e-05,
+      "loss": 4.4114,
+      "step": 201728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.668876276482069e-05,
+      "loss": 4.397,
+      "step": 202240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.668037681731017e-05,
+      "loss": 4.4006,
+      "step": 202752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.667199086979965e-05,
+      "loss": 4.4029,
+      "step": 203264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.666360492228913e-05,
+      "loss": 4.4032,
+      "step": 203776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6655235353582346e-05,
+      "loss": 4.403,
+      "step": 204288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.664684940607182e-05,
+      "loss": 4.3936,
+      "step": 204800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.66384634585613e-05,
+      "loss": 4.3938,
+      "step": 205312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.663007751105078e-05,
+      "loss": 4.4027,
+      "step": 205824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.662169156354026e-05,
+      "loss": 4.3747,
+      "step": 206336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.661330561602974e-05,
+      "loss": 4.3935,
+      "step": 206848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.660493604732295e-05,
+      "loss": 4.3854,
+      "step": 207360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.659655009981243e-05,
+      "loss": 4.3997,
+      "step": 207872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.658816415230191e-05,
+      "loss": 4.3887,
+      "step": 208384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.657977820479139e-05,
+      "loss": 4.3962,
+      "step": 208896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.657139225728087e-05,
+      "loss": 4.3898,
+      "step": 209408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6563006309770355e-05,
+      "loss": 4.3851,
+      "step": 209920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6554620362259835e-05,
+      "loss": 4.3973,
+      "step": 210432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6546250793553044e-05,
+      "loss": 4.386,
+      "step": 210944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6537864846042524e-05,
+      "loss": 4.3954,
+      "step": 211456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6529478898532004e-05,
+      "loss": 4.3925,
+      "step": 211968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.652110932982521e-05,
+      "loss": 4.3919,
+      "step": 212480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.651272338231469e-05,
+      "loss": 4.3831,
+      "step": 212992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.650433743480417e-05,
+      "loss": 4.3796,
+      "step": 213504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.3948,
+      "step": 214016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.3772,
+      "step": 214528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.647917959227261e-05,
+      "loss": 4.3946,
+      "step": 215040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.647079364476209e-05,
+      "loss": 4.3787,
+      "step": 215552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.646240769725157e-05,
+      "loss": 4.3835,
+      "step": 216064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.645403812854479e-05,
+      "loss": 4.38,
+      "step": 216576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.644565218103427e-05,
+      "loss": 4.3778,
+      "step": 217088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.643726623352375e-05,
+      "loss": 4.3607,
+      "step": 217600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.642888028601323e-05,
+      "loss": 4.3779,
+      "step": 218112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.642051071730644e-05,
+      "loss": 4.3655,
+      "step": 218624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.641212476979592e-05,
+      "loss": 4.3798,
+      "step": 219136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.64037388222854e-05,
+      "loss": 4.3927,
+      "step": 219648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.639535287477488e-05,
+      "loss": 4.3891,
+      "step": 220160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.638696692726436e-05,
+      "loss": 4.3567,
+      "step": 220672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6378597358557566e-05,
+      "loss": 4.3821,
+      "step": 221184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6370211411047046e-05,
+      "loss": 4.3757,
+      "step": 221696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6361825463536526e-05,
+      "loss": 4.3753,
+      "step": 222208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6353439516026006e-05,
+      "loss": 4.3666,
+      "step": 222720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6345053568515486e-05,
+      "loss": 4.3765,
+      "step": 223232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6336667621004966e-05,
+      "loss": 4.3706,
+      "step": 223744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6328281673494446e-05,
+      "loss": 4.3682,
+      "step": 224256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6319912104787655e-05,
+      "loss": 4.3717,
+      "step": 224768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6311526157277135e-05,
+      "loss": 4.3739,
+      "step": 225280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6303140209766615e-05,
+      "loss": 4.3692,
+      "step": 225792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6294754262256095e-05,
+      "loss": 4.3661,
+      "step": 226304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6286368314745575e-05,
+      "loss": 4.371,
+      "step": 226816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6277982367235055e-05,
+      "loss": 4.3722,
+      "step": 227328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6269596419724535e-05,
+      "loss": 4.3629,
+      "step": 227840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6261226851017744e-05,
+      "loss": 4.3769,
+      "step": 228352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.625284090350723e-05,
+      "loss": 4.3624,
+      "step": 228864
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.36259651184082,
+      "eval_runtime": 281.8171,
+      "eval_samples_per_second": 1354.038,
+      "eval_steps_per_second": 42.315,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624445495599671e-05,
+      "loss": 4.3523,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623606900848619e-05,
+      "loss": 4.3554,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622768306097567e-05,
+      "loss": 4.3738,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621929711346515e-05,
+      "loss": 4.3614,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621091116595463e-05,
+      "loss": 4.3739,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620252521844411e-05,
+      "loss": 4.3485,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619413927093359e-05,
+      "loss": 4.3677,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618575332342307e-05,
+      "loss": 4.3376,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6177367375912543e-05,
+      "loss": 4.3689,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6168981428402023e-05,
+      "loss": 4.3499,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61605954808915e-05,
+      "loss": 4.3641,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615220953338098e-05,
+      "loss": 4.363,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61438399646742e-05,
+      "loss": 4.3491,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613545401716368e-05,
+      "loss": 4.3518,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612706806965316e-05,
+      "loss": 4.3458,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611868212214264e-05,
+      "loss": 4.3414,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611029617463212e-05,
+      "loss": 4.352,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61019102271216e-05,
+      "loss": 4.3498,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.609352427961108e-05,
+      "loss": 4.3521,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.608513833210056e-05,
+      "loss": 4.3614,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607675238459004e-05,
+      "loss": 4.3643,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606836643707952e-05,
+      "loss": 4.3536,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6059980489569e-05,
+      "loss": 4.3458,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605159454205848e-05,
+      "loss": 4.359,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604322497335169e-05,
+      "loss": 4.3428,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60348554046449e-05,
+      "loss": 4.3568,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602648583593811e-05,
+      "loss": 4.351,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601809988842759e-05,
+      "loss": 4.3422,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600971394091707e-05,
+      "loss": 4.3384,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600132799340655e-05,
+      "loss": 4.3436,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599294204589603e-05,
+      "loss": 4.3393,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598455609838551e-05,
+      "loss": 4.3479,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597617015087499e-05,
+      "loss": 4.3482,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596778420336447e-05,
+      "loss": 4.3443,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595939825585395e-05,
+      "loss": 4.3463,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595101230834343e-05,
+      "loss": 4.3437,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594264273963664e-05,
+      "loss": 4.3266,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.3339,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3267,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3326,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3329,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590071300208404e-05,
+      "loss": 4.3378,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5892359812180986e-05,
+      "loss": 4.3337,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5883973864670466e-05,
+      "loss": 4.3265,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5875587917159946e-05,
+      "loss": 4.3394,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5867201969649426e-05,
+      "loss": 4.3375,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5858816022138906e-05,
+      "loss": 4.3404,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585043007462838e-05,
+      "loss": 4.3315,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584204412711786e-05,
+      "loss": 4.3203,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5833674558411075e-05,
+      "loss": 4.336,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5825288610900555e-05,
+      "loss": 4.323,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5816902663390035e-05,
+      "loss": 4.3331,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5808516715879515e-05,
+      "loss": 4.3224,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.322,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3222,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3276,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.3248,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3249,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3239,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5749831462109604e-05,
+      "loss": 4.3161,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3129,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5733059567088564e-05,
+      "loss": 4.3364,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724673619578044e-05,
+      "loss": 4.3026,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716287672067524e-05,
+      "loss": 4.3051,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5707901724557004e-05,
+      "loss": 4.311,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699515777046484e-05,
+      "loss": 4.3285,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691129829535963e-05,
+      "loss": 4.3135,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568276026082918e-05,
+      "loss": 4.31,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.567437431331866e-05,
+      "loss": 4.3038,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.566598836580814e-05,
+      "loss": 4.3108,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565760241829762e-05,
+      "loss": 4.3221,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564923284959083e-05,
+      "loss": 4.3159,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3144,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.3249,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5624091385863e-05,
+      "loss": 4.3232,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561570543835248e-05,
+      "loss": 4.3053,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560731949084196e-05,
+      "loss": 4.3058,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559893354333144e-05,
+      "loss": 4.3199,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559054759582092e-05,
+      "loss": 4.2946,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55821616483104e-05,
+      "loss": 4.3137,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.3014,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556540613209309e-05,
+      "loss": 4.3239,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557020184582566e-05,
+      "loss": 4.3087,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548634237072046e-05,
+      "loss": 4.297,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540248289561526e-05,
+      "loss": 4.297,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.311,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.3169,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5515106825833695e-05,
+      "loss": 4.3058,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5506720878323175e-05,
+      "loss": 4.3044,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5498334930812655e-05,
+      "loss": 4.2982,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548994898330214e-05,
+      "loss": 4.3082,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548156303579162e-05,
+      "loss": 4.2984,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54731770882811e-05,
+      "loss": 4.298,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546479114077058e-05,
+      "loss": 4.3085,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.3071,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3035,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.2908,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.2998,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.2959,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541449183451119e-05,
+      "loss": 4.3019,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.540610588700067e-05,
+      "loss": 4.2984,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539775269709761e-05,
+      "loss": 4.2895,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389366749587095e-05,
+      "loss": 4.2926,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5380980802076575e-05,
+      "loss": 4.2985,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372594854566055e-05,
+      "loss": 4.2734,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5364208907055535e-05,
+      "loss": 4.2933,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5355839338348744e-05,
+      "loss": 4.28,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5347453390838224e-05,
+      "loss": 4.2975,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5339067443327704e-05,
+      "loss": 4.288,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5330681495817184e-05,
+      "loss": 4.2913,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5322295548306664e-05,
+      "loss": 4.2864,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5313909600796144e-05,
+      "loss": 4.2876,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5305523653285624e-05,
+      "loss": 4.2979,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5297137705775104e-05,
+      "loss": 4.2815,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528878451587205e-05,
+      "loss": 4.2949,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528039856836153e-05,
+      "loss": 4.2935,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527201262085101e-05,
+      "loss": 4.2891,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526362667334049e-05,
+      "loss": 4.292,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.525524072582997e-05,
+      "loss": 4.277,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524685477831945e-05,
+      "loss": 4.297,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523846883080893e-05,
+      "loss": 4.2828,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52300828832984e-05,
+      "loss": 4.2912,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522171331459162e-05,
+      "loss": 4.2811,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52133273670811e-05,
+      "loss": 4.2863,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5204974177178036e-05,
+      "loss": 4.2831,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5196588229667516e-05,
+      "loss": 4.278,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5188202282157e-05,
+      "loss": 4.267,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517981633464648e-05,
+      "loss": 4.2776,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517143038713596e-05,
+      "loss": 4.2729,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516304443962544e-05,
+      "loss": 4.2836,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515465849211492e-05,
+      "loss": 4.295,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51462725446044e-05,
+      "loss": 4.2923,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5137886597093876e-05,
+      "loss": 4.2642,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5129500649583356e-05,
+      "loss": 4.2892,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5121114702072836e-05,
+      "loss": 4.2801,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5112728754562316e-05,
+      "loss": 4.2813,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5104359185855525e-05,
+      "loss": 4.2735,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5095973238345005e-05,
+      "loss": 4.2833,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5087587290834485e-05,
+      "loss": 4.2729,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507920134332397e-05,
+      "loss": 4.281,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507083177461718e-05,
+      "loss": 4.2749,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5062462205910396e-05,
+      "loss": 4.278,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5054076258399876e-05,
+      "loss": 4.2787,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504569031088935e-05,
+      "loss": 4.272,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503730436337883e-05,
+      "loss": 4.2771,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502891841586831e-05,
+      "loss": 4.2835,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502054884716152e-05,
+      "loss": 4.2678,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5012162899651e-05,
+      "loss": 4.2813,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500377695214048e-05,
+      "loss": 4.2755,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.2788496017456055,
+      "eval_runtime": 304.6964,
+      "eval_samples_per_second": 1252.365,
+      "eval_steps_per_second": 39.137,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.499539100462996e-05,
+      "loss": 4.2668,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.498700505711944e-05,
+      "loss": 4.2638,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978619109608925e-05,
+      "loss": 4.2771,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4970233162098405e-05,
+      "loss": 4.2719,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4961847214587885e-05,
+      "loss": 4.2869,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4953477645881094e-05,
+      "loss": 4.2568,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49451080771743e-05,
+      "loss": 4.2738,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.493672212966378e-05,
+      "loss": 4.2555,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492833618215326e-05,
+      "loss": 4.2732,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491996661344647e-05,
+      "loss": 4.2622,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491158066593595e-05,
+      "loss": 4.2721,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490319471842543e-05,
+      "loss": 4.2766,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489480877091491e-05,
+      "loss": 4.2586,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.488642282340439e-05,
+      "loss": 4.2673,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487803687589387e-05,
+      "loss": 4.2572,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486965092838336e-05,
+      "loss": 4.2514,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486126498087284e-05,
+      "loss": 4.2623,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485287903336232e-05,
+      "loss": 4.2615,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.48444930858518e-05,
+      "loss": 4.2654,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.483610713834128e-05,
+      "loss": 4.274,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.482772119083076e-05,
+      "loss": 4.2784,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4819368000927697e-05,
+      "loss": 4.2673,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4810982053417177e-05,
+      "loss": 4.2596,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4802596105906656e-05,
+      "loss": 4.2743,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4794210158396136e-05,
+      "loss": 4.2542,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4785824210885616e-05,
+      "loss": 4.2697,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4777438263375096e-05,
+      "loss": 4.2664,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4769052315864576e-05,
+      "loss": 4.2551,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4760666368354056e-05,
+      "loss": 4.2533,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4752280420843536e-05,
+      "loss": 4.2558,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474392723094048e-05,
+      "loss": 4.2558,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473554128342996e-05,
+      "loss": 4.2616,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472715533591944e-05,
+      "loss": 4.2645,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471876938840892e-05,
+      "loss": 4.2612,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.47103834408984e-05,
+      "loss": 4.2613,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470201387219161e-05,
+      "loss": 4.2607,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469362792468109e-05,
+      "loss": 4.2418,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468524197717057e-05,
+      "loss": 4.2541,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467685602966005e-05,
+      "loss": 4.2415,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466847008214953e-05,
+      "loss": 4.249,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4660100513442746e-05,
+      "loss": 4.2486,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4651714565932226e-05,
+      "loss": 4.2607,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4643328618421706e-05,
+      "loss": 4.2501,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.463494267091118e-05,
+      "loss": 4.2439,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.462655672340066e-05,
+      "loss": 4.2592,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618187154693875e-05,
+      "loss": 4.2516,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609801207183355e-05,
+      "loss": 4.2565,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460141525967283e-05,
+      "loss": 4.2536,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.459302931216231e-05,
+      "loss": 4.2381,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.458464336465179e-05,
+      "loss": 4.2557,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576273795945004e-05,
+      "loss": 4.2424,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4567887848434484e-05,
+      "loss": 4.2535,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4559501900923964e-05,
+      "loss": 4.2416,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4551115953413444e-05,
+      "loss": 4.2403,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.454274638470665e-05,
+      "loss": 4.2429,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453436043719613e-05,
+      "loss": 4.2466,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452597448968561e-05,
+      "loss": 4.2428,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451758854217509e-05,
+      "loss": 4.2448,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450920259466457e-05,
+      "loss": 4.2459,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450081664715405e-05,
+      "loss": 4.2362,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.449243069964353e-05,
+      "loss": 4.2316,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.448404475213301e-05,
+      "loss": 4.2608,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.447565880462249e-05,
+      "loss": 4.2216,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446727285711197e-05,
+      "loss": 4.226,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445891966720892e-05,
+      "loss": 4.2347,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44505337196984e-05,
+      "loss": 4.2525,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.444214777218788e-05,
+      "loss": 4.23,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.443376182467736e-05,
+      "loss": 4.2373,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4425392255970566e-05,
+      "loss": 4.225,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4417006308460046e-05,
+      "loss": 4.234,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4408620360949526e-05,
+      "loss": 4.2421,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4400234413439006e-05,
+      "loss": 4.2394,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4391848465928486e-05,
+      "loss": 4.2381,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4383462518417966e-05,
+      "loss": 4.2448,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4375076570907446e-05,
+      "loss": 4.2446,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4366690623396926e-05,
+      "loss": 4.2321,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435832105469014e-05,
+      "loss": 4.2282,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434993510717962e-05,
+      "loss": 4.2445,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43415491596691e-05,
+      "loss": 4.2196,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433316321215858e-05,
+      "loss": 4.2383,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432477726464806e-05,
+      "loss": 4.2258,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431639131713754e-05,
+      "loss": 4.2482,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.430802174843075e-05,
+      "loss": 4.2345,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429963580092023e-05,
+      "loss": 4.2241,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429124985340971e-05,
+      "loss": 4.2179,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428286390589919e-05,
+      "loss": 4.2386,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4274477958388664e-05,
+      "loss": 4.2388,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4266092010878144e-05,
+      "loss": 4.2362,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.425772244217136e-05,
+      "loss": 4.2286,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424933649466084e-05,
+      "loss": 4.2263,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424095054715032e-05,
+      "loss": 4.2328,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.42325645996398e-05,
+      "loss": 4.2228,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422417865212928e-05,
+      "loss": 4.2242,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.421579270461876e-05,
+      "loss": 4.2351,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.420740675710824e-05,
+      "loss": 4.2317,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419902080959772e-05,
+      "loss": 4.2299,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419065124089093e-05,
+      "loss": 4.2196,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.418226529338041e-05,
+      "loss": 4.2283,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.417387934586989e-05,
+      "loss": 4.219,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.416549339835937e-05,
+      "loss": 4.2285,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.415710745084885e-05,
+      "loss": 4.2248,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414873788214206e-05,
+      "loss": 4.2216,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414036831343527e-05,
+      "loss": 4.2185,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.413198236592475e-05,
+      "loss": 4.2195,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.412359641841423e-05,
+      "loss": 4.2126,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411521047090371e-05,
+      "loss": 4.2167,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.410684090219692e-05,
+      "loss": 4.2088,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40984549546864e-05,
+      "loss": 4.2263,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409006900717588e-05,
+      "loss": 4.2197,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408168305966536e-05,
+      "loss": 4.2174,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407329711215484e-05,
+      "loss": 4.2145,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406491116464432e-05,
+      "loss": 4.2226,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40565252171338e-05,
+      "loss": 4.2234,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404813926962328e-05,
+      "loss": 4.2105,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40397697009165e-05,
+      "loss": 4.2245,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403140013220971e-05,
+      "loss": 4.2259,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.402301418469919e-05,
+      "loss": 4.2188,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401462823718867e-05,
+      "loss": 4.2233,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4006242289678147e-05,
+      "loss": 4.207,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3997856342167627e-05,
+      "loss": 4.223,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3989470394657106e-05,
+      "loss": 4.2179,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3981084447146586e-05,
+      "loss": 4.2242,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3972698499636066e-05,
+      "loss": 4.2047,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3964312552125546e-05,
+      "loss": 4.2206,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3955942983418755e-05,
+      "loss": 4.2165,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3947557035908235e-05,
+      "loss": 4.209,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393918746720145e-05,
+      "loss": 4.2008,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393080151969093e-05,
+      "loss": 4.2097,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.392241557218041e-05,
+      "loss": 4.2048,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.391402962466989e-05,
+      "loss": 4.219,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390564367715937e-05,
+      "loss": 4.2209,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389725772964885e-05,
+      "loss": 4.2267,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3888871782138324e-05,
+      "loss": 4.1969,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388050221343154e-05,
+      "loss": 4.2213,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.387211626592102e-05,
+      "loss": 4.2157,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38637303184105e-05,
+      "loss": 4.2083,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.385534437089997e-05,
+      "loss": 4.2068,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.384695842338945e-05,
+      "loss": 4.22,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383857247587893e-05,
+      "loss": 4.2079,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383018652836842e-05,
+      "loss": 4.2134,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.382181695966163e-05,
+      "loss": 4.2043,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381343101215111e-05,
+      "loss": 4.2101,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.380504506464059e-05,
+      "loss": 4.2115,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379665911713007e-05,
+      "loss": 4.2083,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378827316961955e-05,
+      "loss": 4.2104,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377988722210903e-05,
+      "loss": 4.217,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377151765340224e-05,
+      "loss": 4.2032,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.376313170589172e-05,
+      "loss": 4.2122,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.37547457583812e-05,
+      "loss": 4.2123,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.22148323059082,
+      "eval_runtime": 308.9825,
+      "eval_samples_per_second": 1234.992,
+      "eval_steps_per_second": 38.594,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374635981087068e-05,
+      "loss": 4.2003,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373797386336016e-05,
+      "loss": 4.1975,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372958791584964e-05,
+      "loss": 4.2126,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372120196833912e-05,
+      "loss": 4.2083,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37128160208286e-05,
+      "loss": 4.2157,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704430073318084e-05,
+      "loss": 4.2004,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3696044125807564e-05,
+      "loss": 4.2073,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368765817829704e-05,
+      "loss": 4.1919,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367927223078652e-05,
+      "loss": 4.206,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3670886283276e-05,
+      "loss": 4.2027,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366250033576548e-05,
+      "loss": 4.2067,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365411438825496e-05,
+      "loss": 4.2097,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364572844074444e-05,
+      "loss": 4.1987,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363734249323392e-05,
+      "loss": 4.2005,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36289565457234e-05,
+      "loss": 4.197,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362057059821288e-05,
+      "loss": 4.1888,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361218465070236e-05,
+      "loss": 4.1955,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360379870319184e-05,
+      "loss": 4.201,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359542913448505e-05,
+      "loss": 4.2019,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358704318697453e-05,
+      "loss": 4.2132,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357865723946401e-05,
+      "loss": 4.2146,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357027129195349e-05,
+      "loss": 4.2077,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356188534444297e-05,
+      "loss": 4.1989,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355349939693245e-05,
+      "loss": 4.2108,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354512982822566e-05,
+      "loss": 4.1954,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353674388071514e-05,
+      "loss": 4.2069,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352835793320462e-05,
+      "loss": 4.2016,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35199719856941e-05,
+      "loss": 4.1914,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511586038183575e-05,
+      "loss": 4.1923,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503200090673055e-05,
+      "loss": 4.1987,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3494814143162535e-05,
+      "loss": 4.1952,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348642819565202e-05,
+      "loss": 4.199,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347805862694523e-05,
+      "loss": 4.2045,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346967267943471e-05,
+      "loss": 4.1982,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461303110727926e-05,
+      "loss": 4.199,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34529171632174e-05,
+      "loss": 4.2007,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344453121570688e-05,
+      "loss": 4.1864,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343614526819636e-05,
+      "loss": 4.1867,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3427775699489575e-05,
+      "loss": 4.1829,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341938975197905e-05,
+      "loss": 4.1879,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341100380446853e-05,
+      "loss": 4.1881,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340261785695801e-05,
+      "loss": 4.2036,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339423190944749e-05,
+      "loss": 4.1907,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385845961936975e-05,
+      "loss": 4.1835,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377460014426455e-05,
+      "loss": 4.2015,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369074066915935e-05,
+      "loss": 4.1872,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360704498209144e-05,
+      "loss": 4.2005,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352318550698624e-05,
+      "loss": 4.1905,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3343932603188104e-05,
+      "loss": 4.1802,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333556303448131e-05,
+      "loss": 4.196,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332717708697079e-05,
+      "loss": 4.1851,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331879113946027e-05,
+      "loss": 4.1975,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331040519194975e-05,
+      "loss": 4.1792,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330203562324296e-05,
+      "loss": 4.185,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329364967573244e-05,
+      "loss": 4.1827,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328526372822193e-05,
+      "loss": 4.1878,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327687778071141e-05,
+      "loss": 4.1888,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326849183320089e-05,
+      "loss": 4.1815,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326010588569037e-05,
+      "loss": 4.1911,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325171993817985e-05,
+      "loss": 4.182,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324333399066933e-05,
+      "loss": 4.1724,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323494804315881e-05,
+      "loss": 4.2052,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322657847445202e-05,
+      "loss": 4.1662,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32181925269415e-05,
+      "loss": 4.1645,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320980657943098e-05,
+      "loss": 4.1782,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320142063192046e-05,
+      "loss": 4.1986,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3193051063213666e-05,
+      "loss": 4.1697,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3184665115703146e-05,
+      "loss": 4.1825,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3176279168192626e-05,
+      "loss": 4.1685,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3167893220682106e-05,
+      "loss": 4.1764,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3159507273171586e-05,
+      "loss": 4.186,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315115408326853e-05,
+      "loss": 4.1849,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314276813575801e-05,
+      "loss": 4.1833,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313438218824749e-05,
+      "loss": 4.1843,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312599624073697e-05,
+      "loss": 4.1923,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311761029322645e-05,
+      "loss": 4.1727,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310922434571593e-05,
+      "loss": 4.1764,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310085477700914e-05,
+      "loss": 4.1851,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309246882949862e-05,
+      "loss": 4.1653,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30840828819881e-05,
+      "loss": 4.1772,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307569693447758e-05,
+      "loss": 4.1754,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306731098696706e-05,
+      "loss": 4.188,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3058941418260276e-05,
+      "loss": 4.1815,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050555470749756e-05,
+      "loss": 4.17,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042169523239236e-05,
+      "loss": 4.1653,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303378357572871e-05,
+      "loss": 4.1808,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302539762821819e-05,
+      "loss": 4.181,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301701168070767e-05,
+      "loss": 4.1774,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300862573319715e-05,
+      "loss": 4.1795,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300023978568663e-05,
+      "loss": 4.1719,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299187021697984e-05,
+      "loss": 4.1774,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298348426946932e-05,
+      "loss": 4.1707,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29750983219588e-05,
+      "loss": 4.1694,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2966712374448285e-05,
+      "loss": 4.1837,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2958342805741494e-05,
+      "loss": 4.1798,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949956858230974e-05,
+      "loss": 4.1749,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2941570910720454e-05,
+      "loss": 4.1651,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293320134201366e-05,
+      "loss": 4.1759,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292481539450314e-05,
+      "loss": 4.1623,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291642944699262e-05,
+      "loss": 4.1766,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29080434994821e-05,
+      "loss": 4.1721,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289965755197158e-05,
+      "loss": 4.1697,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289127160446106e-05,
+      "loss": 4.169,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288288565695054e-05,
+      "loss": 4.1588,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287449970944002e-05,
+      "loss": 4.1661,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286613014073324e-05,
+      "loss": 4.16,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285776057202645e-05,
+      "loss": 4.1611,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284937462451593e-05,
+      "loss": 4.1696,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284098867700541e-05,
+      "loss": 4.1691,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283260272949489e-05,
+      "loss": 4.1645,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282421678198437e-05,
+      "loss": 4.1592,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281583083447385e-05,
+      "loss": 4.1678,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807461265767056e-05,
+      "loss": 4.1725,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799075318256536e-05,
+      "loss": 4.1613,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2790689370746016e-05,
+      "loss": 4.1744,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2782303423235496e-05,
+      "loss": 4.1719,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2773917475724976e-05,
+      "loss": 4.1698,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276554790701819e-05,
+      "loss": 4.1685,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275716195950767e-05,
+      "loss": 4.1591,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274877601199715e-05,
+      "loss": 4.1654,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274039006448663e-05,
+      "loss": 4.1686,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273202049577984e-05,
+      "loss": 4.1714,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272363454826932e-05,
+      "loss": 4.1572,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27152486007588e-05,
+      "loss": 4.1665,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270686265324828e-05,
+      "loss": 4.1661,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269847670573776e-05,
+      "loss": 4.1581,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269009075822724e-05,
+      "loss": 4.1471,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2681704810716714e-05,
+      "loss": 4.1622,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673318863206194e-05,
+      "loss": 4.1522,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266494929449941e-05,
+      "loss": 4.1675,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265656334698889e-05,
+      "loss": 4.1708,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264817739947837e-05,
+      "loss": 4.1743,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263979145196785e-05,
+      "loss": 4.148,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2631421883261065e-05,
+      "loss": 4.1704,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2623052314554274e-05,
+      "loss": 4.166,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614666367043754e-05,
+      "loss": 4.1582,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606280419533234e-05,
+      "loss": 4.1554,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597894472022714e-05,
+      "loss": 4.1711,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258950852451219e-05,
+      "loss": 4.1539,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258112257700167e-05,
+      "loss": 4.1668,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257273662949115e-05,
+      "loss": 4.1529,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256435068198063e-05,
+      "loss": 4.1631,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255598111327384e-05,
+      "loss": 4.1605,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254761154456706e-05,
+      "loss": 4.1572,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253922559705654e-05,
+      "loss": 4.1638,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253083964954602e-05,
+      "loss": 4.1593,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252245370203549e-05,
+      "loss": 4.1615,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251406775452497e-05,
+      "loss": 4.1555,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250569818581819e-05,
+      "loss": 4.1669,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1800079345703125,
+      "eval_runtime": 293.153,
+      "eval_samples_per_second": 1301.679,
+      "eval_steps_per_second": 40.678,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249731223830766e-05,
+      "loss": 4.1548,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248892629079714e-05,
+      "loss": 4.1462,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248054034328662e-05,
+      "loss": 4.1604,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247217077457984e-05,
+      "loss": 4.1628,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246378482706932e-05,
+      "loss": 4.1667,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24553988795588e-05,
+      "loss": 4.1523,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244701293204828e-05,
+      "loss": 4.1576,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243862698453776e-05,
+      "loss": 4.1432,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2430257415830966e-05,
+      "loss": 4.1591,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2421871468320446e-05,
+      "loss": 4.151,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2413485520809926e-05,
+      "loss": 4.1615,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2405099573299406e-05,
+      "loss": 4.1596,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2396730004592615e-05,
+      "loss": 4.1506,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2388344057082095e-05,
+      "loss": 4.1559,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2379958109571575e-05,
+      "loss": 4.1442,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2371572162061055e-05,
+      "loss": 4.1408,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2363186214550535e-05,
+      "loss": 4.1476,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2354800267040015e-05,
+      "loss": 4.1571,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23464143195295e-05,
+      "loss": 4.1497,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233802837201898e-05,
+      "loss": 4.1681,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232964242450846e-05,
+      "loss": 4.1623,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232125647699794e-05,
+      "loss": 4.1617,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231288690829115e-05,
+      "loss": 4.1513,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230450096078063e-05,
+      "loss": 4.1613,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229611501327011e-05,
+      "loss": 4.1457,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228772906575959e-05,
+      "loss": 4.1614,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227934311824907e-05,
+      "loss": 4.1515,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227095717073855e-05,
+      "loss": 4.1451,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226257122322802e-05,
+      "loss": 4.15,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.22541852757175e-05,
+      "loss": 4.1495,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2245832085814455e-05,
+      "loss": 4.1485,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2237446138303935e-05,
+      "loss": 4.1562,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2229060190793415e-05,
+      "loss": 4.1534,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2220674243282895e-05,
+      "loss": 4.157,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2212304674576104e-05,
+      "loss": 4.1449,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2203918727065584e-05,
+      "loss": 4.153,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2195532779555064e-05,
+      "loss": 4.1473,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2187146832044544e-05,
+      "loss": 4.1349,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2178760884534024e-05,
+      "loss": 4.1371,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21703749370235e-05,
+      "loss": 4.1414,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216198898951298e-05,
+      "loss": 4.1431,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215360304200246e-05,
+      "loss": 4.1549,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214521709449194e-05,
+      "loss": 4.1444,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213684752578515e-05,
+      "loss": 4.1379,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212846157827463e-05,
+      "loss": 4.1563,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212007563076411e-05,
+      "loss": 4.1417,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211170606205732e-05,
+      "loss": 4.1548,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210333649335054e-05,
+      "loss": 4.1456,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209495054584002e-05,
+      "loss": 4.1278,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20865645983295e-05,
+      "loss": 4.1579,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207817865081897e-05,
+      "loss": 4.1357,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206979270330845e-05,
+      "loss": 4.1551,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061423134601667e-05,
+      "loss": 4.1371,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2053037187091147e-05,
+      "loss": 4.1376,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2044651239580626e-05,
+      "loss": 4.1331,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036265292070106e-05,
+      "loss": 4.141,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2027879344559586e-05,
+      "loss": 4.1485,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019493397049066e-05,
+      "loss": 4.1328,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011107449538546e-05,
+      "loss": 4.1476,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2002737880831755e-05,
+      "loss": 4.1406,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1994351933321235e-05,
+      "loss": 4.1299,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1985965985810715e-05,
+      "loss": 4.1552,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1977580038300195e-05,
+      "loss": 4.125,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1969194090789675e-05,
+      "loss": 4.1206,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1960808143279155e-05,
+      "loss": 4.1369,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1952422195768635e-05,
+      "loss": 4.1481,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1944036248258115e-05,
+      "loss": 4.1263,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1935666679551324e-05,
+      "loss": 4.1418,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192728073204081e-05,
+      "loss": 4.1214,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191891116333402e-05,
+      "loss": 4.1377,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19105252158235e-05,
+      "loss": 4.1399,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.190213926831298e-05,
+      "loss": 4.1372,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.189375332080246e-05,
+      "loss": 4.1392,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188536737329194e-05,
+      "loss": 4.1386,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187698142578142e-05,
+      "loss": 4.1503,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186861185707463e-05,
+      "loss": 4.1292,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186022590956411e-05,
+      "loss": 4.1358,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.185183996205359e-05,
+      "loss": 4.1359,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184345401454307e-05,
+      "loss": 4.1261,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.183506806703255e-05,
+      "loss": 4.1327,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182668211952203e-05,
+      "loss": 4.1312,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181829617201151e-05,
+      "loss": 4.1441,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180991022450099e-05,
+      "loss": 4.1423,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1801540655794204e-05,
+      "loss": 4.1202,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1793154708283684e-05,
+      "loss": 4.1235,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1784785139576894e-05,
+      "loss": 4.1365,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776399192066373e-05,
+      "loss": 4.1411,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1768013244555853e-05,
+      "loss": 4.1354,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.175962729704533e-05,
+      "loss": 4.1305,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1751241349534807e-05,
+      "loss": 4.1338,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1742855402024287e-05,
+      "loss": 4.1327,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1734469454513766e-05,
+      "loss": 4.1311,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1726083507003246e-05,
+      "loss": 4.1259,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171769755949273e-05,
+      "loss": 4.1383,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170932799078594e-05,
+      "loss": 4.1349,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170095842207916e-05,
+      "loss": 4.1352,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169257247456863e-05,
+      "loss": 4.1233,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168418652705811e-05,
+      "loss": 4.1307,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167580057954759e-05,
+      "loss": 4.1252,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166741463203707e-05,
+      "loss": 4.1343,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165904506333028e-05,
+      "loss": 4.1293,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165065911581976e-05,
+      "loss": 4.1237,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164227316830924e-05,
+      "loss": 4.1305,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163388722079872e-05,
+      "loss": 4.1178,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16255012732882e-05,
+      "loss": 4.1232,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161711532577769e-05,
+      "loss": 4.1179,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1608745757070896e-05,
+      "loss": 4.1221,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1600359809560376e-05,
+      "loss": 4.1207,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1591973862049856e-05,
+      "loss": 4.1349,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1583587914539336e-05,
+      "loss": 4.118,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1575201967028816e-05,
+      "loss": 4.1211,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1566816019518296e-05,
+      "loss": 4.1228,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1558446450811505e-05,
+      "loss": 4.1304,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1550060503300985e-05,
+      "loss": 4.1208,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1541674555790465e-05,
+      "loss": 4.1361,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1533288608279945e-05,
+      "loss": 4.1282,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1524902660769425e-05,
+      "loss": 4.1258,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151653309206264e-05,
+      "loss": 4.128,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150814714455212e-05,
+      "loss": 4.1221,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14997611970416e-05,
+      "loss": 4.1252,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149137524953108e-05,
+      "loss": 4.1297,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.148298930202056e-05,
+      "loss": 4.1251,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147460335451004e-05,
+      "loss": 4.1188,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146621740699952e-05,
+      "loss": 4.1288,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145784783829273e-05,
+      "loss": 4.1257,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144946189078221e-05,
+      "loss": 4.1214,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144107594327169e-05,
+      "loss": 4.1042,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143268999576117e-05,
+      "loss": 4.119,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142430404825064e-05,
+      "loss": 4.1105,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141591810074012e-05,
+      "loss": 4.1274,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14075321532296e-05,
+      "loss": 4.1299,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139914620571909e-05,
+      "loss": 4.1318,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13907766370123e-05,
+      "loss": 4.1087,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138242344710924e-05,
+      "loss": 4.1291,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.137403749959872e-05,
+      "loss": 4.1242,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13656515520882e-05,
+      "loss": 4.1227,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.135726560457768e-05,
+      "loss": 4.1168,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134887965706716e-05,
+      "loss": 4.1298,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134049370955664e-05,
+      "loss": 4.113,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1332107762046116e-05,
+      "loss": 4.1298,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1323721814535596e-05,
+      "loss": 4.1092,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131535224582881e-05,
+      "loss": 4.1244,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130696629831829e-05,
+      "loss": 4.1215,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129858035080777e-05,
+      "loss": 4.1181,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129019440329725e-05,
+      "loss": 4.1232,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128182483459047e-05,
+      "loss": 4.1198,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127343888707994e-05,
+      "loss": 4.1213,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126505293956942e-05,
+      "loss": 4.1163,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12566669920589e-05,
+      "loss": 4.1297,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.14900016784668,
+      "eval_runtime": 294.7575,
+      "eval_samples_per_second": 1294.593,
+      "eval_steps_per_second": 40.457,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124828104454838e-05,
+      "loss": 4.1181,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123989509703786e-05,
+      "loss": 4.1063,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123150914952734e-05,
+      "loss": 4.1246,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122312320201682e-05,
+      "loss": 4.1221,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12147372545063e-05,
+      "loss": 4.1254,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120635130699578e-05,
+      "loss": 4.1144,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119796535948526e-05,
+      "loss": 4.1177,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118957941197474e-05,
+      "loss": 4.1066,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118119346446423e-05,
+      "loss": 4.1187,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117280751695371e-05,
+      "loss": 4.1158,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116442156944318e-05,
+      "loss": 4.119,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115603562193266e-05,
+      "loss": 4.122,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114764967442214e-05,
+      "loss": 4.1171,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113926372691162e-05,
+      "loss": 4.112,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11308777794011e-05,
+      "loss": 4.1085,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112249183189058e-05,
+      "loss": 4.1011,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111410588438006e-05,
+      "loss": 4.1108,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110571993686954e-05,
+      "loss": 4.1178,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109733398935902e-05,
+      "loss": 4.1122,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10889480418485e-05,
+      "loss": 4.1243,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108057847314171e-05,
+      "loss": 4.128,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072192525631196e-05,
+      "loss": 4.1261,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1063806578120676e-05,
+      "loss": 4.1145,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1055437009413885e-05,
+      "loss": 4.1177,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1047051061903365e-05,
+      "loss": 4.1137,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1038665114392845e-05,
+      "loss": 4.1213,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030279166882325e-05,
+      "loss": 4.1121,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1021893219371805e-05,
+      "loss": 4.1123,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1013507271861285e-05,
+      "loss": 4.1129,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1005121324350764e-05,
+      "loss": 4.1099,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0996735376840244e-05,
+      "loss": 4.1074,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0988365808133454e-05,
+      "loss": 4.1176,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0979979860622933e-05,
+      "loss": 4.1176,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0971593913112413e-05,
+      "loss": 4.1194,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096322434440563e-05,
+      "loss": 4.1062,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095483839689511e-05,
+      "loss": 4.1205,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094645244938459e-05,
+      "loss": 4.1127,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093806650187407e-05,
+      "loss": 4.0952,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092968055436354e-05,
+      "loss": 4.1003,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092129460685302e-05,
+      "loss": 4.1032,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09129086593425e-05,
+      "loss": 4.109,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.090453909063572e-05,
+      "loss": 4.115,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089615314312519e-05,
+      "loss": 4.1102,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088776719561467e-05,
+      "loss": 4.0979,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087938124810415e-05,
+      "loss": 4.1219,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087099530059363e-05,
+      "loss": 4.1035,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086260935308311e-05,
+      "loss": 4.1196,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08542234055726e-05,
+      "loss": 4.1119,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084585383686581e-05,
+      "loss": 4.0844,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0837484268159016e-05,
+      "loss": 4.125,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0829098320648496e-05,
+      "loss": 4.0974,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0820712373137976e-05,
+      "loss": 4.1151,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0812326425627456e-05,
+      "loss": 4.1034,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0803940478116936e-05,
+      "loss": 4.1,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0795554530606416e-05,
+      "loss": 4.1004,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0787168583095896e-05,
+      "loss": 4.0991,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0778782635585376e-05,
+      "loss": 4.1171,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0770396688074856e-05,
+      "loss": 4.0926,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0762027119368065e-05,
+      "loss": 4.1095,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075365755066128e-05,
+      "loss": 4.1088,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074527160315076e-05,
+      "loss": 4.0905,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073688565564024e-05,
+      "loss": 4.116,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072849970812972e-05,
+      "loss": 4.094,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07201137606192e-05,
+      "loss": 4.0885,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071172781310868e-05,
+      "loss": 4.0975,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070334186559816e-05,
+      "loss": 4.1108,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069495591808764e-05,
+      "loss": 4.0908,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068656997057712e-05,
+      "loss": 4.1061,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06781840230666e-05,
+      "loss": 4.0878,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066979807555608e-05,
+      "loss": 4.0982,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0661412128045553e-05,
+      "loss": 4.1069,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065304255933877e-05,
+      "loss": 4.1012,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064465661182825e-05,
+      "loss": 4.1065,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063627066431773e-05,
+      "loss": 4.1006,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062788471680721e-05,
+      "loss": 4.1119,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061949876929669e-05,
+      "loss": 4.0975,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061111282178617e-05,
+      "loss": 4.1023,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060274325307938e-05,
+      "loss": 4.0951,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059435730556886e-05,
+      "loss": 4.0909,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058597135805834e-05,
+      "loss": 4.0978,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057758541054782e-05,
+      "loss": 4.0988,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056923222064476e-05,
+      "loss": 4.1067,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056084627313424e-05,
+      "loss": 4.1123,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055246032562372e-05,
+      "loss": 4.0825,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05440743781132e-05,
+      "loss": 4.0874,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053568843060268e-05,
+      "loss": 4.102,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05273188618959e-05,
+      "loss": 4.1066,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051893291438538e-05,
+      "loss": 4.1006,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051054696687485e-05,
+      "loss": 4.0951,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050216101936433e-05,
+      "loss": 4.1036,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049377507185381e-05,
+      "loss": 4.0952,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048538912434329e-05,
+      "loss": 4.0981,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04770195556365e-05,
+      "loss": 4.0903,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046863360812598e-05,
+      "loss": 4.1003,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046024766061546e-05,
+      "loss": 4.1026,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045186171310494e-05,
+      "loss": 4.1019,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044347576559443e-05,
+      "loss": 4.0891,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043508981808391e-05,
+      "loss": 4.0916,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042670387057339e-05,
+      "loss": 4.0884,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041831792306287e-05,
+      "loss": 4.0999,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0409948354356076e-05,
+      "loss": 4.0985,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0401562406845556e-05,
+      "loss": 4.0884,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0393176459335036e-05,
+      "loss": 4.0952,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0384790511824516e-05,
+      "loss": 4.0854,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0376420943117725e-05,
+      "loss": 4.0857,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0368034995607205e-05,
+      "loss": 4.0816,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0359665426900414e-05,
+      "loss": 4.0892,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0351279479389894e-05,
+      "loss": 4.0891,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0342893531879374e-05,
+      "loss": 4.0981,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033450758436886e-05,
+      "loss": 4.0885,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032612163685834e-05,
+      "loss": 4.0803,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031773568934782e-05,
+      "loss": 4.0901,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03093497418373e-05,
+      "loss": 4.0927,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030098017313051e-05,
+      "loss": 4.0913,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029259422561999e-05,
+      "loss": 4.0969,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028420827810947e-05,
+      "loss": 4.093,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027582233059895e-05,
+      "loss": 4.1002,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026743638308843e-05,
+      "loss": 4.0919,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025905043557791e-05,
+      "loss": 4.0879,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025066448806738e-05,
+      "loss": 4.0883,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024227854055686e-05,
+      "loss": 4.1008,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.023390897185008e-05,
+      "loss": 4.0886,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022552302433956e-05,
+      "loss": 4.0823,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021713707682904e-05,
+      "loss": 4.0933,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020875112931852e-05,
+      "loss": 4.0971,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0200381560611735e-05,
+      "loss": 4.0852,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0191995613101214e-05,
+      "loss": 4.0761,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0183626044394424e-05,
+      "loss": 4.0843,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175240096883904e-05,
+      "loss": 4.0779,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0166854149373383e-05,
+      "loss": 4.0933,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015846820186286e-05,
+      "loss": 4.0971,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015008225435234e-05,
+      "loss": 4.0991,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0141696306841817e-05,
+      "loss": 4.0732,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013332673813503e-05,
+      "loss": 4.098,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012494079062451e-05,
+      "loss": 4.0907,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011655484311399e-05,
+      "loss": 4.0889,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010816889560347e-05,
+      "loss": 4.0837,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009978294809295e-05,
+      "loss": 4.0948,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009139700058243e-05,
+      "loss": 4.0823,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008301105307191e-05,
+      "loss": 4.0987,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007464148436512e-05,
+      "loss": 4.0788,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00662555368546e-05,
+      "loss": 4.0885,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005786958934408e-05,
+      "loss": 4.0882,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004948364183356e-05,
+      "loss": 4.087,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004109769432304e-05,
+      "loss": 4.0894,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003271174681252e-05,
+      "loss": 4.0867,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0024325799302e-05,
+      "loss": 4.0928,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001593985179148e-05,
+      "loss": 4.0855,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00075702830847e-05,
+      "loss": 4.093,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.123981952667236,
+      "eval_runtime": 292.8994,
+      "eval_samples_per_second": 1302.805,
+      "eval_steps_per_second": 40.714,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999918433557418e-05,
+      "loss": 4.0813,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999079838806366e-05,
+      "loss": 4.0783,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998241244055314e-05,
+      "loss": 4.0879,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997402649304262e-05,
+      "loss": 4.0884,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99656405455321e-05,
+      "loss": 4.0946,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957254598021577e-05,
+      "loss": 4.0819,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994886865051105e-05,
+      "loss": 4.0902,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994048270300053e-05,
+      "loss": 4.077,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993209675549001e-05,
+      "loss": 4.0814,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992371080797949e-05,
+      "loss": 4.0854,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991532486046897e-05,
+      "loss": 4.0872,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990693891295845e-05,
+      "loss": 4.0885,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989855296544793e-05,
+      "loss": 4.0887,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890167017937416e-05,
+      "loss": 4.0766,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9881781070426896e-05,
+      "loss": 4.077,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873395122916376e-05,
+      "loss": 4.0699,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865025554209585e-05,
+      "loss": 4.0803,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9856639606699065e-05,
+      "loss": 4.0907,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848253659188545e-05,
+      "loss": 4.077,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9839867711678025e-05,
+      "loss": 4.0925,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9831481764167505e-05,
+      "loss": 4.0951,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823095816656985e-05,
+      "loss": 4.0935,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981470986914646e-05,
+      "loss": 4.0889,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980632392163594e-05,
+      "loss": 4.0813,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9797954352929154e-05,
+      "loss": 4.0823,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9789568405418634e-05,
+      "loss": 4.0868,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781182457908114e-05,
+      "loss": 4.0875,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977281288920133e-05,
+      "loss": 4.081,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976442694169081e-05,
+      "loss": 4.0831,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975604099418029e-05,
+      "loss": 4.079,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974765504666976e-05,
+      "loss": 4.0742,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973926909915924e-05,
+      "loss": 4.0859,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973088315164872e-05,
+      "loss": 4.084,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97224972041382e-05,
+      "loss": 4.0888,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971411125662768e-05,
+      "loss": 4.0766,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970574168792089e-05,
+      "loss": 4.0873,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969735574041037e-05,
+      "loss": 4.0841,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968896979289985e-05,
+      "loss": 4.0635,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968058384538934e-05,
+      "loss": 4.0736,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967219789787882e-05,
+      "loss": 4.0703,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96638119503683e-05,
+      "loss": 4.0753,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965542600285778e-05,
+      "loss": 4.0886,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964704005534726e-05,
+      "loss": 4.0784,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963867048664047e-05,
+      "loss": 4.0714,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963028453912995e-05,
+      "loss": 4.0852,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9621914970423156e-05,
+      "loss": 4.0749,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613529022912636e-05,
+      "loss": 4.0896,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605143075402116e-05,
+      "loss": 4.0809,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9596757127891596e-05,
+      "loss": 4.0558,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9588371180381076e-05,
+      "loss": 4.0937,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9579985232870556e-05,
+      "loss": 4.0688,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9571599285360036e-05,
+      "loss": 4.0849,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956322971665325e-05,
+      "loss": 4.0741,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955484376914273e-05,
+      "loss": 4.07,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954645782163221e-05,
+      "loss": 4.0701,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953807187412169e-05,
+      "loss": 4.0696,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952968592661117e-05,
+      "loss": 4.0895,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952129997910065e-05,
+      "loss": 4.0598,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9512914031590125e-05,
+      "loss": 4.0838,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9504528084079605e-05,
+      "loss": 4.076,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496142136569085e-05,
+      "loss": 4.058,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9487772567862294e-05,
+      "loss": 4.0865,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479386620351774e-05,
+      "loss": 4.0663,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947100067284126e-05,
+      "loss": 4.0543,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946261472533074e-05,
+      "loss": 4.0677,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945422877782022e-05,
+      "loss": 4.0815,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445875587917166e-05,
+      "loss": 4.0651,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437489640406646e-05,
+      "loss": 4.0753,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429103692896126e-05,
+      "loss": 4.0627,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94207177453856e-05,
+      "loss": 4.0655,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9412348176678815e-05,
+      "loss": 4.0767,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9403962229168295e-05,
+      "loss": 4.0741,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939557628165777e-05,
+      "loss": 4.0753,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938719033414725e-05,
+      "loss": 4.0721,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937880438663673e-05,
+      "loss": 4.0852,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370418439126214e-05,
+      "loss": 4.0663,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362032491615694e-05,
+      "loss": 4.0727,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9353646544105174e-05,
+      "loss": 4.0668,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934527697539838e-05,
+      "loss": 4.0648,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933690740669159e-05,
+      "loss": 4.0709,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932852145918107e-05,
+      "loss": 4.0664,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932013551167055e-05,
+      "loss": 4.0777,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931174956416003e-05,
+      "loss": 4.0859,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930336361664951e-05,
+      "loss": 4.0546,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929497766913899e-05,
+      "loss": 4.059,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928659172162847e-05,
+      "loss": 4.0729,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927820577411795e-05,
+      "loss": 4.0738,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926983620541116e-05,
+      "loss": 4.0735,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926145025790065e-05,
+      "loss": 4.0684,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925306431039013e-05,
+      "loss": 4.0724,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924467836287961e-05,
+      "loss": 4.0671,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923630879417282e-05,
+      "loss": 4.0664,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92279228466623e-05,
+      "loss": 4.0651,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921953689915178e-05,
+      "loss": 4.0728,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921115095164126e-05,
+      "loss": 4.0748,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9202781382934466e-05,
+      "loss": 4.0735,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9194395435423946e-05,
+      "loss": 4.0583,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9186009487913426e-05,
+      "loss": 4.0679,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9177623540402906e-05,
+      "loss": 4.0604,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169237592892386e-05,
+      "loss": 4.07,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9160851645381866e-05,
+      "loss": 4.0714,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9152465697871346e-05,
+      "loss": 4.0575,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914407975036083e-05,
+      "loss": 4.068,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913569380285031e-05,
+      "loss": 4.0537,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912732423414352e-05,
+      "loss": 4.0605,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9118938286633e-05,
+      "loss": 4.0524,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911056871792621e-05,
+      "loss": 4.0661,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910218277041569e-05,
+      "loss": 4.0615,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909379682290517e-05,
+      "loss": 4.0677,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908541087539465e-05,
+      "loss": 4.0572,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907702492788413e-05,
+      "loss": 4.0569,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068638980373604e-05,
+      "loss": 4.0588,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060253032863083e-05,
+      "loss": 4.0697,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905186708535257e-05,
+      "loss": 4.062,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9043513895449515e-05,
+      "loss": 4.0656,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035127947938995e-05,
+      "loss": 4.0656,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026742000428475e-05,
+      "loss": 4.0736,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018356052917955e-05,
+      "loss": 4.0655,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900997010540743e-05,
+      "loss": 4.0593,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900158415789691e-05,
+      "loss": 4.0635,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899319821038639e-05,
+      "loss": 4.0709,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898481226287587e-05,
+      "loss": 4.0621,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897642631536535e-05,
+      "loss": 4.0573,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896805674665856e-05,
+      "loss": 4.0628,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895967079914804e-05,
+      "loss": 4.0692,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8951284851637524e-05,
+      "loss": 4.0583,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8942898904127004e-05,
+      "loss": 4.0475,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893452933542021e-05,
+      "loss": 4.0554,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892614338790969e-05,
+      "loss": 4.0508,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891775744039917e-05,
+      "loss": 4.0689,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890937149288865e-05,
+      "loss": 4.0672,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890100192418186e-05,
+      "loss": 4.0716,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889261597667134e-05,
+      "loss": 4.0482,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888423002916082e-05,
+      "loss": 4.073,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88758440816503e-05,
+      "loss": 4.0637,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886747451294351e-05,
+      "loss": 4.0641,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885908856543299e-05,
+      "loss": 4.0597,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885070261792248e-05,
+      "loss": 4.0613,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884231667041196e-05,
+      "loss": 4.0584,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883393072290144e-05,
+      "loss": 4.0721,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882554477539092e-05,
+      "loss": 4.0501,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88171588278804e-05,
+      "loss": 4.0629,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880877288036988e-05,
+      "loss": 4.0646,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8800403311663086e-05,
+      "loss": 4.0638,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792017364152566e-05,
+      "loss": 4.0571,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8783631416642046e-05,
+      "loss": 4.0614,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8775261847935255e-05,
+      "loss": 4.0653,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8766875900424735e-05,
+      "loss": 4.06,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8758489952914215e-05,
+      "loss": 4.0664,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.104763507843018,
+      "eval_runtime": 282.9882,
+      "eval_samples_per_second": 1348.434,
+      "eval_steps_per_second": 42.14,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750104005403695e-05,
+      "loss": 4.0554,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8741718057893175e-05,
+      "loss": 4.0481,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8733332110382655e-05,
+      "loss": 4.0606,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872494616287214e-05,
+      "loss": 4.0626,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8716560215361615e-05,
+      "loss": 4.0677,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8708174267851095e-05,
+      "loss": 4.0552,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869980469914431e-05,
+      "loss": 4.0635,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869141875163379e-05,
+      "loss": 4.0471,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683032804123264e-05,
+      "loss": 4.0568,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8674646856612744e-05,
+      "loss": 4.0595,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866627728790596e-05,
+      "loss": 4.0605,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865789134039544e-05,
+      "loss": 4.0636,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864952177168865e-05,
+      "loss": 4.0677,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864113582417813e-05,
+      "loss": 4.0484,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863274987666761e-05,
+      "loss": 4.0493,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862436392915709e-05,
+      "loss": 4.0383,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861597798164657e-05,
+      "loss": 4.0571,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860759203413605e-05,
+      "loss": 4.0626,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859920608662553e-05,
+      "loss": 4.0501,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859082013911501e-05,
+      "loss": 4.0637,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858245057040822e-05,
+      "loss": 4.0717,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8574081001701434e-05,
+      "loss": 4.0658,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8565695054190914e-05,
+      "loss": 4.062,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855730910668039e-05,
+      "loss": 4.0552,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85489395379736e-05,
+      "loss": 4.0605,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854055359046308e-05,
+      "loss": 4.0595,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853216764295256e-05,
+      "loss": 4.0621,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852378169544204e-05,
+      "loss": 4.0604,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851539574793152e-05,
+      "loss": 4.0541,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8507009800421e-05,
+      "loss": 4.0489,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849862385291048e-05,
+      "loss": 4.0498,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849023790539996e-05,
+      "loss": 4.0609,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848186833669317e-05,
+      "loss": 4.0622,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847348238918265e-05,
+      "loss": 4.0627,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846509644167213e-05,
+      "loss": 4.0517,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845672687296534e-05,
+      "loss": 4.0547,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448357304258556e-05,
+      "loss": 4.0588,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8439971356748036e-05,
+      "loss": 4.0394,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8431585409237516e-05,
+      "loss": 4.0461,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423199461726996e-05,
+      "loss": 4.0462,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414813514216476e-05,
+      "loss": 4.0475,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8406427566705956e-05,
+      "loss": 4.0634,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8398041619195436e-05,
+      "loss": 4.0536,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8389655671684916e-05,
+      "loss": 4.0467,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8381269724174396e-05,
+      "loss": 4.0607,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8372883776663876e-05,
+      "loss": 4.0469,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8364497829153356e-05,
+      "loss": 4.0597,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8356111881642836e-05,
+      "loss": 4.0643,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8347725934132316e-05,
+      "loss": 4.0298,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8339339986621796e-05,
+      "loss": 4.063,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8330954039111276e-05,
+      "loss": 4.0465,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8322584470404485e-05,
+      "loss": 4.0557,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831419852289397e-05,
+      "loss": 4.0505,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830581257538345e-05,
+      "loss": 4.0425,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8297426627872925e-05,
+      "loss": 4.0456,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828905705916614e-05,
+      "loss": 4.0417,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828067111165562e-05,
+      "loss": 4.0637,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82722851641451e-05,
+      "loss": 4.0343,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8263899216634574e-05,
+      "loss": 4.0599,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8255513269124054e-05,
+      "loss": 4.053,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8247127321613534e-05,
+      "loss": 4.0351,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8238741374103013e-05,
+      "loss": 4.0613,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230355426592493e-05,
+      "loss": 4.0401,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822198585788571e-05,
+      "loss": 4.0302,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821359991037519e-05,
+      "loss": 4.0412,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820521396286467e-05,
+      "loss": 4.0591,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819682801535415e-05,
+      "loss": 4.0367,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818845844664736e-05,
+      "loss": 4.0527,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818007249913684e-05,
+      "loss": 4.0375,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817168655162632e-05,
+      "loss": 4.037,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81633006041158e-05,
+      "loss": 4.0492,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815491465660528e-05,
+      "loss": 4.0497,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814654508789849e-05,
+      "loss": 4.0523,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813815914038797e-05,
+      "loss": 4.0479,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812977319287745e-05,
+      "loss": 4.057,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812140362417066e-05,
+      "loss": 4.0464,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811301767666014e-05,
+      "loss": 4.0483,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810463172914962e-05,
+      "loss": 4.0375,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80962457816391e-05,
+      "loss": 4.0423,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808785983412858e-05,
+      "loss": 4.0469,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807947388661806e-05,
+      "loss": 4.04,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807108793910754e-05,
+      "loss": 4.0538,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806270199159702e-05,
+      "loss": 4.0581,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805434880169396e-05,
+      "loss": 4.0382,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804596285418344e-05,
+      "loss": 4.0307,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803757690667292e-05,
+      "loss": 4.0477,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80291909591624e-05,
+      "loss": 4.0488,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802080501165188e-05,
+      "loss": 4.05,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801241906414136e-05,
+      "loss": 4.0407,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800403311663084e-05,
+      "loss": 4.0498,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799564716912033e-05,
+      "loss": 4.0417,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7987277600413536e-05,
+      "loss": 4.0408,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7978891652903016e-05,
+      "loss": 4.0412,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7970505705392496e-05,
+      "loss": 4.0476,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7962136136685705e-05,
+      "loss": 4.0477,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7953750189175185e-05,
+      "loss": 4.0494,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7945364241664665e-05,
+      "loss": 4.0332,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7936978294154145e-05,
+      "loss": 4.0452,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7928592346643625e-05,
+      "loss": 4.0347,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7920206399133105e-05,
+      "loss": 4.0472,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791182045162258e-05,
+      "loss": 4.0462,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7903434504112065e-05,
+      "loss": 4.0385,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789506493540528e-05,
+      "loss": 4.0434,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788667898789476e-05,
+      "loss": 4.0293,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7878293040384234e-05,
+      "loss": 4.0373,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7869907092873714e-05,
+      "loss": 4.0297,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786153752416693e-05,
+      "loss": 4.0407,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785315157665641e-05,
+      "loss": 4.0385,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784476562914588e-05,
+      "loss": 4.0425,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783637968163536e-05,
+      "loss": 4.0334,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782799373412484e-05,
+      "loss": 4.0343,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781962416541805e-05,
+      "loss": 4.0373,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781123821790753e-05,
+      "loss": 4.0446,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780285227039702e-05,
+      "loss": 4.0387,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7794482701690235e-05,
+      "loss": 4.04,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778609675417971e-05,
+      "loss": 4.0404,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777771080666919e-05,
+      "loss": 4.0481,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776932485915867e-05,
+      "loss": 4.0423,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776093891164815e-05,
+      "loss": 4.0359,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775256934294136e-05,
+      "loss": 4.0403,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774418339543084e-05,
+      "loss": 4.046,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773579744792032e-05,
+      "loss": 4.0413,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77274115004098e-05,
+      "loss": 4.0348,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771902555289928e-05,
+      "loss": 4.0348,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7710639605388757e-05,
+      "loss": 4.0459,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770227003668197e-05,
+      "loss": 4.0328,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769388408917145e-05,
+      "loss": 4.0256,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768549814166093e-05,
+      "loss": 4.0315,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767711219415041e-05,
+      "loss": 4.0254,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766872624663989e-05,
+      "loss": 4.0506,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766034029912937e-05,
+      "loss": 4.0367,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765195435161885e-05,
+      "loss": 4.0516,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764358478291206e-05,
+      "loss": 4.0267,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763519883540154e-05,
+      "loss": 4.0459,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762681288789102e-05,
+      "loss": 4.0385,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76184269403805e-05,
+      "loss": 4.0428,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761004099286998e-05,
+      "loss": 4.0364,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760165504535946e-05,
+      "loss": 4.0364,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759326909784894e-05,
+      "loss": 4.0316,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758488315033842e-05,
+      "loss": 4.0507,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75764972028279e-05,
+      "loss": 4.0269,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756811125531738e-05,
+      "loss": 4.0418,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755972530780686e-05,
+      "loss": 4.0423,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755135573910007e-05,
+      "loss": 4.0368,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754296979158955e-05,
+      "loss": 4.0322,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753458384407903e-05,
+      "loss": 4.039,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752619789656851e-05,
+      "loss": 4.0422,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751782832786172e-05,
+      "loss": 4.04,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75094423803512e-05,
+      "loss": 4.0437,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.08948278427124,
+      "eval_runtime": 284.0045,
+      "eval_samples_per_second": 1343.609,
+      "eval_steps_per_second": 41.989,
+      "step": 763200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.750105643284068e-05,
+      "loss": 4.0267,
+      "step": 763392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.749267048533016e-05,
+      "loss": 4.0255,
+      "step": 763904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7484300916623375e-05,
+      "loss": 4.0386,
+      "step": 764416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7475914969112855e-05,
+      "loss": 4.0396,
+      "step": 764928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7467529021602335e-05,
+      "loss": 4.0476,
+      "step": 765440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7459143074091815e-05,
+      "loss": 4.035,
+      "step": 765952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7450757126581294e-05,
+      "loss": 4.0386,
+      "step": 766464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7442371179070774e-05,
+      "loss": 4.0284,
+      "step": 766976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7433985231560254e-05,
+      "loss": 4.0321,
+      "step": 767488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7425599284049734e-05,
+      "loss": 4.0354,
+      "step": 768000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7417213336539214e-05,
+      "loss": 4.0339,
+      "step": 768512
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7408843767832423e-05,
+      "loss": 4.0428,
+      "step": 769024
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.740047419912563e-05,
+      "loss": 4.0425,
+      "step": 769536
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.739208825161511e-05,
+      "loss": 4.0254,
+      "step": 770048
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.738370230410459e-05,
+      "loss": 4.0278,
+      "step": 770560
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.737531635659407e-05,
+      "loss": 4.0162,
+      "step": 771072
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.736693040908356e-05,
+      "loss": 4.0377,
+      "step": 771584
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.735854446157304e-05,
+      "loss": 4.0364,
+      "step": 772096
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.735015851406252e-05,
+      "loss": 4.0266,
+      "step": 772608
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.7341772566552e-05,
+      "loss": 4.0372,
+      "step": 773120
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.733340299784521e-05,
+      "loss": 4.0532,
+      "step": 773632
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.732501705033469e-05,
+      "loss": 4.0419,
+      "step": 774144
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.73166474816279e-05,
+      "loss": 4.0385,
+      "step": 774656
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.730826153411738e-05,
+      "loss": 4.0339,
+      "step": 775168
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.729987558660686e-05,
+      "loss": 4.0392,
+      "step": 775680
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.729148963909634e-05,
+      "loss": 4.0353,
+      "step": 776192
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.728310369158582e-05,
+      "loss": 4.0353,
+      "step": 776704
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.72747177440753e-05,
+      "loss": 4.0392,
+      "step": 777216
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.726633179656478e-05,
+      "loss": 4.0338,
+      "step": 777728
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.725794584905426e-05,
+      "loss": 4.0297,
+      "step": 778240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.724957628034747e-05,
+      "loss": 4.0296,
+      "step": 778752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.724120671164068e-05,
+      "loss": 4.0373,
+      "step": 779264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.723282076413016e-05,
+      "loss": 4.036,
+      "step": 779776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.722443481661964e-05,
+      "loss": 4.0431,
+      "step": 780288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.721604886910912e-05,
+      "loss": 4.0351,
+      "step": 780800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.720767930040233e-05,
+      "loss": 4.0317,
+      "step": 781312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.719929335289181e-05,
+      "loss": 4.0374,
+      "step": 781824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.719090740538129e-05,
+      "loss": 4.0211,
+      "step": 782336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.718252145787077e-05,
+      "loss": 4.0246,
+      "step": 782848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.717413551036025e-05,
+      "loss": 4.024,
+      "step": 783360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.716574956284973e-05,
+      "loss": 4.0229,
+      "step": 783872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.715736361533921e-05,
+      "loss": 4.0365,
+      "step": 784384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.714897766782869e-05,
+      "loss": 4.0321,
+      "step": 784896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.714059172031817e-05,
+      "loss": 4.0321,
+      "step": 785408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.713220577280765e-05,
+      "loss": 4.0312,
+      "step": 785920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.712381982529713e-05,
+      "loss": 4.0273,
+      "step": 786432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.711545025659034e-05,
+      "loss": 4.0384,
+      "step": 786944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.710706430907982e-05,
+      "loss": 4.0399,
+      "step": 787456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.70986783615693e-05,
+      "loss": 4.0084,
+      "step": 787968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.709030879286251e-05,
+      "loss": 4.0418,
+      "step": 788480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.708192284535199e-05,
+      "loss": 4.0226,
+      "step": 788992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.707353689784147e-05,
+      "loss": 4.0371,
+      "step": 789504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.706515095033095e-05,
+      "loss": 4.0313,
+      "step": 790016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.7056765002820435e-05,
+      "loss": 4.0175,
+      "step": 790528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.7048379055309915e-05,
+      "loss": 4.0277,
+      "step": 791040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.7039993107799395e-05,
+      "loss": 4.0169,
+      "step": 791552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.7031607160288875e-05,
+      "loss": 4.0392,
+      "step": 792064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.7023221212778355e-05,
+      "loss": 4.0178,
+      "step": 792576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.7014835265267835e-05,
+      "loss": 4.0356,
+      "step": 793088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.7006449317757315e-05,
+      "loss": 4.031,
+      "step": 793600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6998063370246795e-05,
+      "loss": 4.0176,
+      "step": 794112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6989693801540004e-05,
+      "loss": 4.0378,
+      "step": 794624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6981307854029484e-05,
+      "loss": 4.0168,
+      "step": 795136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6972921906518964e-05,
+      "loss": 4.0116,
+      "step": 795648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.696453595900844e-05,
+      "loss": 4.0153,
+      "step": 796160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.695615001149792e-05,
+      "loss": 4.0371,
+      "step": 796672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.69477640639874e-05,
+      "loss": 4.0174,
+      "step": 797184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.693939449528061e-05,
+      "loss": 4.035,
+      "step": 797696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.693100854777009e-05,
+      "loss": 4.0099,
+      "step": 798208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.692262260025957e-05,
+      "loss": 4.0198,
+      "step": 798720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.691425303155279e-05,
+      "loss": 4.0279,
+      "step": 799232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6905883462846e-05,
+      "loss": 4.0274,
+      "step": 799744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.689749751533548e-05,
+      "loss": 4.0299,
+      "step": 800256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.688911156782496e-05,
+      "loss": 4.0301,
+      "step": 800768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.688072562031444e-05,
+      "loss": 4.0344,
+      "step": 801280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.687233967280391e-05,
+      "loss": 4.0221,
+      "step": 801792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.686395372529339e-05,
+      "loss": 4.0269,
+      "step": 802304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.685556777778287e-05,
+      "loss": 4.018,
+      "step": 802816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.684718183027235e-05,
+      "loss": 4.0216,
+      "step": 803328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.683879588276184e-05,
+      "loss": 4.0238,
+      "step": 803840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.683040993525132e-05,
+      "loss": 4.016,
+      "step": 804352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.68220239877408e-05,
+      "loss": 4.032,
+      "step": 804864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.681363804023028e-05,
+      "loss": 4.0371,
+      "step": 805376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6805268471523486e-05,
+      "loss": 4.0237,
+      "step": 805888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6796898902816695e-05,
+      "loss": 4.0047,
+      "step": 806400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6788512955306175e-05,
+      "loss": 4.0269,
+      "step": 806912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6780127007795655e-05,
+      "loss": 4.0253,
+      "step": 807424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6771741060285135e-05,
+      "loss": 4.0337,
+      "step": 807936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6763355112774615e-05,
+      "loss": 4.0154,
+      "step": 808448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.6754969165264095e-05,
+      "loss": 4.0313,
+      "step": 808960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6746599596557304e-05,
+      "loss": 4.0208,
+      "step": 809472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.673821364904679e-05,
+      "loss": 4.0211,
+      "step": 809984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.672982770153627e-05,
+      "loss": 4.0187,
+      "step": 810496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.672144175402575e-05,
+      "loss": 4.0277,
+      "step": 811008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.671305580651523e-05,
+      "loss": 4.0269,
+      "step": 811520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.670466985900471e-05,
+      "loss": 4.028,
+      "step": 812032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.669630029029792e-05,
+      "loss": 4.0125,
+      "step": 812544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.66879143427874e-05,
+      "loss": 4.0225,
+      "step": 813056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.667952839527688e-05,
+      "loss": 4.0151,
+      "step": 813568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.667114244776636e-05,
+      "loss": 4.0237,
+      "step": 814080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.666275650025584e-05,
+      "loss": 4.022,
+      "step": 814592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.665437055274532e-05,
+      "loss": 4.0178,
+      "step": 815104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.66459846052348e-05,
+      "loss": 4.0205,
+      "step": 815616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.663759865772427e-05,
+      "loss": 4.0082,
+      "step": 816128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.662921271021376e-05,
+      "loss": 4.0205,
+      "step": 816640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6620843141506975e-05,
+      "loss": 4.006,
+      "step": 817152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6612457193996455e-05,
+      "loss": 4.0196,
+      "step": 817664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.660407124648593e-05,
+      "loss": 4.0131,
+      "step": 818176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.659568529897541e-05,
+      "loss": 4.0247,
+      "step": 818688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6587315730268624e-05,
+      "loss": 4.0132,
+      "step": 819200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.65789297827581e-05,
+      "loss": 4.0116,
+      "step": 819712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.657054383524758e-05,
+      "loss": 4.0146,
+      "step": 820224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.656215788773706e-05,
+      "loss": 4.0252,
+      "step": 820736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.655378831903027e-05,
+      "loss": 4.0187,
+      "step": 821248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6545402371519746e-05,
+      "loss": 4.0186,
+      "step": 821760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6537016424009226e-05,
+      "loss": 4.0235,
+      "step": 822272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.652864685530244e-05,
+      "loss": 4.0286,
+      "step": 822784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.652027728659566e-05,
+      "loss": 4.0192,
+      "step": 823296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.651189133908514e-05,
+      "loss": 4.0172,
+      "step": 823808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.650350539157462e-05,
+      "loss": 4.0191,
+      "step": 824320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.64951194440641e-05,
+      "loss": 4.0255,
+      "step": 824832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.648673349655357e-05,
+      "loss": 4.0189,
+      "step": 825344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.647834754904305e-05,
+      "loss": 4.0147,
+      "step": 825856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.646996160153253e-05,
+      "loss": 4.0157,
+      "step": 826368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.646157565402201e-05,
+      "loss": 4.0253,
+      "step": 826880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.645320608531522e-05,
+      "loss": 4.0104,
+      "step": 827392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.64448201378047e-05,
+      "loss": 4.0108,
+      "step": 827904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6436450569097916e-05,
+      "loss": 4.0043,
+      "step": 828416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6428064621587396e-05,
+      "loss": 4.0086,
+      "step": 828928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6419678674076876e-05,
+      "loss": 4.0256,
+      "step": 829440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6411292726566356e-05,
+      "loss": 4.0204,
+      "step": 829952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6402906779055836e-05,
+      "loss": 4.0329,
+      "step": 830464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6394520831545316e-05,
+      "loss": 4.0047,
+      "step": 830976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6386134884034796e-05,
+      "loss": 4.0234,
+      "step": 831488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6377748936524276e-05,
+      "loss": 4.0173,
+      "step": 832000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6369362989013756e-05,
+      "loss": 4.0223,
+      "step": 832512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6360977041503236e-05,
+      "loss": 4.0159,
+      "step": 833024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6352591093992715e-05,
+      "loss": 4.0136,
+      "step": 833536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6344221525285925e-05,
+      "loss": 4.0174,
+      "step": 834048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6335851956579134e-05,
+      "loss": 4.0306,
+      "step": 834560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.6327466009068614e-05,
+      "loss": 4.0035,
+      "step": 835072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.63190800615581e-05,
+      "loss": 4.0208,
+      "step": 835584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.631069411404758e-05,
+      "loss": 4.0196,
+      "step": 836096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.630230816653706e-05,
+      "loss": 4.0169,
+      "step": 836608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.629392221902654e-05,
+      "loss": 4.012,
+      "step": 837120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.628553627151602e-05,
+      "loss": 4.0199,
+      "step": 837632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.62771503240055e-05,
+      "loss": 4.0177,
+      "step": 838144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.626876437649498e-05,
+      "loss": 4.02,
+      "step": 838656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.626037842898446e-05,
+      "loss": 4.0205,
+      "step": 839168
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.075167179107666,
+      "eval_runtime": 298.8637,
+      "eval_samples_per_second": 1276.806,
+      "eval_steps_per_second": 39.901,
+      "step": 839520
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.3615504229118246e+17,
+  "trial_name": null,
+  "trial_params": null
+}