diff --git "a/checkpoint-2518560/trainer_state.json" "b/checkpoint-2518560/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2518560/trainer_state.json"
@@ -0,0 +1,29803 @@
+{
+  "best_metric": 3.9933865070343018,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/lstm/2/checkpoints/checkpoint-2518560",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 2518560,
+  "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
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625199248147393e-05,
+      "loss": 3.9992,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624360653396341e-05,
+      "loss": 4.0075,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623522058645289e-05,
+      "loss": 4.017,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622683463894237e-05,
+      "loss": 4.0182,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621844869143185e-05,
+      "loss": 4.0263,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621006274392133e-05,
+      "loss": 4.0157,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620167679641081e-05,
+      "loss": 4.021,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.61932908489003e-05,
+      "loss": 4.013,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618490490138978e-05,
+      "loss": 4.0087,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.617651895387926e-05,
+      "loss": 4.0109,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.616813300636874e-05,
+      "loss": 4.0176,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615974705885822e-05,
+      "loss": 4.0179,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615137749015143e-05,
+      "loss": 4.0256,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.614299154264091e-05,
+      "loss": 4.0063,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613460559513039e-05,
+      "loss": 4.0082,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612621964761987e-05,
+      "loss": 3.996,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611783370010935e-05,
+      "loss": 4.0175,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610944775259882e-05,
+      "loss": 4.0161,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.61010618050883e-05,
+      "loss": 4.0046,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609267585757778e-05,
+      "loss": 4.0175,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608428991006727e-05,
+      "loss": 4.035,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607590396255675e-05,
+      "loss": 4.0205,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606751801504623e-05,
+      "loss": 4.0178,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605913206753571e-05,
+      "loss": 4.0149,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605076249882892e-05,
+      "loss": 4.019,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6042392930122126e-05,
+      "loss": 4.0167,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6034006982611606e-05,
+      "loss": 4.013,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6025621035101086e-05,
+      "loss": 4.0154,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6017235087590566e-05,
+      "loss": 4.0139,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6008849140080046e-05,
+      "loss": 4.009,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6000463192569526e-05,
+      "loss": 4.0128,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5992077245059006e-05,
+      "loss": 4.0129,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5983691297548486e-05,
+      "loss": 4.0143,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59753217288417e-05,
+      "loss": 4.0253,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596693578133118e-05,
+      "loss": 4.0135,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595858259142812e-05,
+      "loss": 4.0098,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59501966439176e-05,
+      "loss": 4.0202,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594181069640708e-05,
+      "loss": 4.0017,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593342474889656e-05,
+      "loss": 4.0074,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592503880138604e-05,
+      "loss": 4.0015,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591665285387552e-05,
+      "loss": 4.0056,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5908266906365e-05,
+      "loss": 4.011,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589988095885448e-05,
+      "loss": 4.0129,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589149501134396e-05,
+      "loss": 4.0149,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588312544263717e-05,
+      "loss": 4.0108,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5874739495126656e-05,
+      "loss": 4.0123,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5866353547616136e-05,
+      "loss": 4.0161,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5857967600105615e-05,
+      "loss": 4.0192,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5849581652595095e-05,
+      "loss": 3.9917,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5841195705084575e-05,
+      "loss": 4.0219,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5832809757574055e-05,
+      "loss": 4.0022,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5824423810063535e-05,
+      "loss": 4.0149,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581603786255301e-05,
+      "loss": 4.0107,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5807684672649953e-05,
+      "loss": 4.0042,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5799298725139433e-05,
+      "loss": 4.001,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579091277762891e-05,
+      "loss": 3.9973,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578252683011839e-05,
+      "loss": 4.025,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.577414088260787e-05,
+      "loss": 3.9963,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576575493509735e-05,
+      "loss": 4.0174,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575736898758683e-05,
+      "loss": 4.0069,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574898304007631e-05,
+      "loss": 4.0008,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574061347136953e-05,
+      "loss": 4.0154,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573224390266274e-05,
+      "loss": 4.0005,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572385795515222e-05,
+      "loss": 3.9909,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57154720076417e-05,
+      "loss": 3.9964,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.570708606013118e-05,
+      "loss": 4.0186,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569870011262066e-05,
+      "loss": 3.9985,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569031416511013e-05,
+      "loss": 4.0133,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.568192821759961e-05,
+      "loss": 3.9891,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567354227008909e-05,
+      "loss": 4.0014,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566517270138231e-05,
+      "loss": 4.0078,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565678675387179e-05,
+      "loss": 4.0109,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564840080636127e-05,
+      "loss": 4.0086,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564001485885075e-05,
+      "loss": 4.0113,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5631645290143956e-05,
+      "loss": 4.0161,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5623259342633436e-05,
+      "loss": 4.0041,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5614873395122916e-05,
+      "loss": 4.0047,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5606487447612396e-05,
+      "loss": 3.9996,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5598101500101876e-05,
+      "loss": 4.0071,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5589715552591356e-05,
+      "loss": 3.9997,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5581329605080836e-05,
+      "loss": 3.9989,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5572943657570316e-05,
+      "loss": 4.0122,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5564557710059796e-05,
+      "loss": 4.0147,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5556171762549275e-05,
+      "loss": 4.0043,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554780219384249e-05,
+      "loss": 3.9836,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553941624633197e-05,
+      "loss": 4.0083,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553103029882145e-05,
+      "loss": 4.0074,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552264435131093e-05,
+      "loss": 4.0161,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.551427478260414e-05,
+      "loss": 3.9968,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550590521389735e-05,
+      "loss": 4.0094,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549751926638683e-05,
+      "loss": 4.0106,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548913331887631e-05,
+      "loss": 3.9979,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548074737136579e-05,
+      "loss": 3.9979,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547236142385527e-05,
+      "loss": 4.0104,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5463991855148485e-05,
+      "loss": 4.005,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5455605907637965e-05,
+      "loss": 4.0124,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5447219960127445e-05,
+      "loss": 3.9979,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5438834012616925e-05,
+      "loss": 4.0028,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5430464443910134e-05,
+      "loss": 3.9952,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5422078496399614e-05,
+      "loss": 4.0085,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5413692548889094e-05,
+      "loss": 4.0068,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5405306601378574e-05,
+      "loss": 3.9968,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5396920653868054e-05,
+      "loss": 4.0025,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5388534706357534e-05,
+      "loss": 3.992,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5380148758847014e-05,
+      "loss": 4.0005,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537177919014022e-05,
+      "loss": 3.9862,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53633932426297e-05,
+      "loss": 4.0013,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.535500729511918e-05,
+      "loss": 3.9933,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534662134760866e-05,
+      "loss": 4.0031,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533823540009814e-05,
+      "loss": 3.9971,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532984945258762e-05,
+      "loss": 4.0024,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53214635050771e-05,
+      "loss": 3.9897,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531309393637032e-05,
+      "loss": 4.008,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530470798885979e-05,
+      "loss": 3.9978,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529633842015301e-05,
+      "loss": 3.9999,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528795247264249e-05,
+      "loss": 4.002,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527956652513197e-05,
+      "loss": 4.0142,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527118057762144e-05,
+      "loss": 3.9989,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.526279463011092e-05,
+      "loss": 3.9976,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52544086826004e-05,
+      "loss": 4.0021,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524602273508989e-05,
+      "loss": 4.0072,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523763678757937e-05,
+      "loss": 4.0004,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522925084006885e-05,
+      "loss": 4.0022,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5220881271362056e-05,
+      "loss": 3.9947,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5212511702655265e-05,
+      "loss": 4.0055,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5204125755144745e-05,
+      "loss": 3.9913,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5195739807634225e-05,
+      "loss": 3.994,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5187353860123705e-05,
+      "loss": 3.9875,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5178967912613185e-05,
+      "loss": 3.9961,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5170581965102665e-05,
+      "loss": 4.0043,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5162196017592145e-05,
+      "loss": 4.0019,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5153810070081625e-05,
+      "loss": 4.0113,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.514544050137484e-05,
+      "loss": 3.9911,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513705455386432e-05,
+      "loss": 4.0034,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512868498515753e-05,
+      "loss": 4.0046,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512029903764701e-05,
+      "loss": 4.0012,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511191309013649e-05,
+      "loss": 3.9977,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510352714262597e-05,
+      "loss": 3.9923,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509514119511545e-05,
+      "loss": 4.0027,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508675524760493e-05,
+      "loss": 4.0073,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507836930009441e-05,
+      "loss": 3.9943,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506998335258389e-05,
+      "loss": 4.0019,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506159740507337e-05,
+      "loss": 4.0032,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505322783636658e-05,
+      "loss": 3.9999,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504484188885606e-05,
+      "loss": 3.9949,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503645594134554e-05,
+      "loss": 4.0004,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5028069993835025e-05,
+      "loss": 3.9984,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5019716803931964e-05,
+      "loss": 4.0013,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501134723522517e-05,
+      "loss": 4.0071,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.064940929412842,
+      "eval_runtime": 296.824,
+      "eval_samples_per_second": 1285.58,
+      "eval_steps_per_second": 40.175,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.500296128771465e-05,
+      "loss": 3.9923,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.499457534020413e-05,
+      "loss": 3.9858,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.498618939269361e-05,
+      "loss": 3.9987,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.497780344518309e-05,
+      "loss": 4.0037,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.496941749767257e-05,
+      "loss": 4.0048,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.496103155016205e-05,
+      "loss": 4.0008,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.495264560265153e-05,
+      "loss": 4.0005,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.494425965514101e-05,
+      "loss": 3.9971,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.493587370763049e-05,
+      "loss": 3.9903,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492748776011998e-05,
+      "loss": 3.9958,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491911819141319e-05,
+      "loss": 4.0022,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491073224390267e-05,
+      "loss": 3.9986,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.490236267519588e-05,
+      "loss": 4.0109,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.489397672768536e-05,
+      "loss": 3.9862,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488559078017484e-05,
+      "loss": 3.9914,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487720483266432e-05,
+      "loss": 3.9798,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48688188851538e-05,
+      "loss": 4.0004,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486043293764327e-05,
+      "loss": 4.0001,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4852063368936486e-05,
+      "loss": 3.9935,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4843677421425966e-05,
+      "loss": 3.994,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4835291473915446e-05,
+      "loss": 4.0151,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4826905526404926e-05,
+      "loss": 4.0039,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4818519578894406e-05,
+      "loss": 4.0013,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481015001018762e-05,
+      "loss": 3.9958,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48017640626771e-05,
+      "loss": 4.0075,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4793378115166575e-05,
+      "loss": 3.9972,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4784992167656055e-05,
+      "loss": 3.9918,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4776606220145535e-05,
+      "loss": 4.004,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4768220272635015e-05,
+      "loss": 3.9922,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4759834325124495e-05,
+      "loss": 3.9945,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4751448377613975e-05,
+      "loss": 3.9967,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4743062430103455e-05,
+      "loss": 3.9929,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4734692861396664e-05,
+      "loss": 3.9985,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.472630691388615e-05,
+      "loss": 4.0088,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471792096637563e-05,
+      "loss": 3.9932,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470955139766884e-05,
+      "loss": 3.9962,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470116545015832e-05,
+      "loss": 4.0025,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46927795026478e-05,
+      "loss": 3.9825,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468439355513728e-05,
+      "loss": 3.988,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467600760762676e-05,
+      "loss": 3.9899,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466762166011624e-05,
+      "loss": 3.9864,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465923571260572e-05,
+      "loss": 3.9923,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46508497650952e-05,
+      "loss": 3.9976,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464248019638841e-05,
+      "loss": 3.9948,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463409424887789e-05,
+      "loss": 3.9945,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462570830136737e-05,
+      "loss": 3.9936,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4617338732660584e-05,
+      "loss": 4.0006,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4608952785150064e-05,
+      "loss": 4.0013,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4600566837639544e-05,
+      "loss": 3.9801,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4592180890129024e-05,
+      "loss": 4.002,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4583794942618504e-05,
+      "loss": 3.9848,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4575408995107984e-05,
+      "loss": 3.9965,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.456702304759746e-05,
+      "loss": 3.9951,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455863710008694e-05,
+      "loss": 3.9853,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455025115257642e-05,
+      "loss": 3.9853,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45418652050659e-05,
+      "loss": 3.9791,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.453347925755538e-05,
+      "loss": 4.0118,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.452509331004486e-05,
+      "loss": 3.9788,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.451670736253434e-05,
+      "loss": 3.998,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450833779382755e-05,
+      "loss": 3.9909,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449995184631703e-05,
+      "loss": 3.9844,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449156589880651e-05,
+      "loss": 3.9953,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448317995129599e-05,
+      "loss": 3.9878,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.447479400378547e-05,
+      "loss": 3.9758,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.446642443507868e-05,
+      "loss": 3.973,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445803848756816e-05,
+      "loss": 4.0075,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444965254005764e-05,
+      "loss": 3.9833,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444126659254712e-05,
+      "loss": 3.9921,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443289702384033e-05,
+      "loss": 3.9763,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442451107632981e-05,
+      "loss": 3.9865,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.441612512881929e-05,
+      "loss": 3.9861,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440773918130877e-05,
+      "loss": 3.9958,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439935323379826e-05,
+      "loss": 3.993,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4390983665091466e-05,
+      "loss": 3.9945,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4382597717580946e-05,
+      "loss": 3.9981,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4374211770070426e-05,
+      "loss": 3.9859,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4365825822559906e-05,
+      "loss": 3.9888,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4357472632656844e-05,
+      "loss": 3.9833,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4349086685146324e-05,
+      "loss": 3.9888,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4340700737635804e-05,
+      "loss": 3.9879,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4332314790125284e-05,
+      "loss": 3.9797,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4323928842614764e-05,
+      "loss": 3.9984,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.431555927390798e-05,
+      "loss": 4.0006,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.430717332639746e-05,
+      "loss": 3.9898,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429878737888694e-05,
+      "loss": 3.9643,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429040143137642e-05,
+      "loss": 3.9889,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.42820154838659e-05,
+      "loss": 3.9918,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.427362953635538e-05,
+      "loss": 3.9979,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.426524358884486e-05,
+      "loss": 3.9875,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.425685764133434e-05,
+      "loss": 3.9904,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424848807262755e-05,
+      "loss": 3.9925,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424010212511703e-05,
+      "loss": 3.982,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.423173255641024e-05,
+      "loss": 3.983,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422334660889972e-05,
+      "loss": 3.9931,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.42149606613892e-05,
+      "loss": 3.9848,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4206591092682414e-05,
+      "loss": 3.9983,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4198205145171894e-05,
+      "loss": 3.9852,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4189819197661374e-05,
+      "loss": 3.9829,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4181433250150854e-05,
+      "loss": 3.9784,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4173047302640333e-05,
+      "loss": 3.9911,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4164661355129813e-05,
+      "loss": 3.9924,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4156275407619293e-05,
+      "loss": 3.979,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4147889460108767e-05,
+      "loss": 3.9851,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4139503512598247e-05,
+      "loss": 3.9758,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4131117565087726e-05,
+      "loss": 3.9891,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4122731617577206e-05,
+      "loss": 3.9661,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4114362048870416e-05,
+      "loss": 3.9833,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4105976101359895e-05,
+      "loss": 3.9823,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409759015384938e-05,
+      "loss": 3.9886,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408920420633886e-05,
+      "loss": 3.9765,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408081825882834e-05,
+      "loss": 3.987,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.407243231131782e-05,
+      "loss": 3.9761,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40640463638073e-05,
+      "loss": 3.9912,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.405566041629678e-05,
+      "loss": 3.9821,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.404729084758999e-05,
+      "loss": 3.9834,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40389212788832e-05,
+      "loss": 3.9837,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403053533137268e-05,
+      "loss": 3.9999,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.402216576266589e-05,
+      "loss": 3.9836,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.401377981515537e-05,
+      "loss": 3.9817,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.400539386764485e-05,
+      "loss": 3.9867,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3997007920134336e-05,
+      "loss": 3.9868,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3988621972623816e-05,
+      "loss": 3.9822,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3980236025113296e-05,
+      "loss": 3.9849,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3971850077602776e-05,
+      "loss": 3.9812,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3963464130092256e-05,
+      "loss": 3.9908,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3955078182581736e-05,
+      "loss": 3.972,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3946708613874945e-05,
+      "loss": 3.9838,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3938322666364425e-05,
+      "loss": 3.9732,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3929936718853905e-05,
+      "loss": 3.9814,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3921550771343385e-05,
+      "loss": 3.9813,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3913181202636594e-05,
+      "loss": 3.9888,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3904795255126074e-05,
+      "loss": 3.9955,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3896409307615554e-05,
+      "loss": 3.9771,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3888023360105034e-05,
+      "loss": 3.9867,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387963741259452e-05,
+      "loss": 3.9849,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3871251465084e-05,
+      "loss": 3.9855,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.386289827518094e-05,
+      "loss": 3.9851,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.385451232767042e-05,
+      "loss": 3.975,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.38461263801599e-05,
+      "loss": 3.9888,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.383774043264938e-05,
+      "loss": 3.9928,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382935448513886e-05,
+      "loss": 3.9769,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382096853762834e-05,
+      "loss": 3.9854,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.381258259011782e-05,
+      "loss": 3.9907,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.38041966426073e-05,
+      "loss": 3.9793,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.379581069509678e-05,
+      "loss": 3.9817,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.378742474758626e-05,
+      "loss": 3.9868,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377903880007574e-05,
+      "loss": 3.9815,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377065285256522e-05,
+      "loss": 3.9824,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.376228328385843e-05,
+      "loss": 3.993,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.054635524749756,
+      "eval_runtime": 297.0083,
+      "eval_samples_per_second": 1284.782,
+      "eval_steps_per_second": 40.15,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375389733634791e-05,
+      "loss": 3.9884,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374551138883739e-05,
+      "loss": 3.9675,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373712544132687e-05,
+      "loss": 3.983,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372873949381635e-05,
+      "loss": 3.9911,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372035354630583e-05,
+      "loss": 3.9848,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371196759879531e-05,
+      "loss": 3.9884,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370358165128479e-05,
+      "loss": 3.9812,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369519570377427e-05,
+      "loss": 3.9833,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368680975626375e-05,
+      "loss": 3.9739,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367842380875323e-05,
+      "loss": 3.9806,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367003786124271e-05,
+      "loss": 3.9858,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3661651913732187e-05,
+      "loss": 3.984,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36532823450254e-05,
+      "loss": 3.9952,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.364489639751488e-05,
+      "loss": 3.9701,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.363651045000436e-05,
+      "loss": 3.9749,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362812450249384e-05,
+      "loss": 3.9641,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3619738554983316e-05,
+      "loss": 3.9853,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3611352607472795e-05,
+      "loss": 3.9819,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3602966659962275e-05,
+      "loss": 3.9775,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3594580712451755e-05,
+      "loss": 3.9784,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3586194764941235e-05,
+      "loss": 3.998,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3577825196234444e-05,
+      "loss": 3.9877,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3569439248723924e-05,
+      "loss": 3.9848,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3561053301213404e-05,
+      "loss": 3.9815,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355266735370289e-05,
+      "loss": 3.9962,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354428140619237e-05,
+      "loss": 3.9792,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353589545868185e-05,
+      "loss": 3.9817,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352750951117133e-05,
+      "loss": 3.9866,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351913994246454e-05,
+      "loss": 3.9763,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351077037375775e-05,
+      "loss": 3.979,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350238442624723e-05,
+      "loss": 3.9812,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349399847873671e-05,
+      "loss": 3.9762,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348561253122619e-05,
+      "loss": 3.9867,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347722658371567e-05,
+      "loss": 3.9915,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346884063620515e-05,
+      "loss": 3.9803,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346047106749836e-05,
+      "loss": 3.9804,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3452085119987845e-05,
+      "loss": 3.9925,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3443699172477325e-05,
+      "loss": 3.962,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3435313224966805e-05,
+      "loss": 3.9755,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3426927277456285e-05,
+      "loss": 3.9776,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3418541329945765e-05,
+      "loss": 3.9672,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3410155382435245e-05,
+      "loss": 3.9788,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3401769434924724e-05,
+      "loss": 3.9857,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3393383487414204e-05,
+      "loss": 3.9804,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338499753990368e-05,
+      "loss": 3.9796,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337661159239316e-05,
+      "loss": 3.9752,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336822564488264e-05,
+      "loss": 3.9821,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3359856076175853e-05,
+      "loss": 3.9878,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3351470128665327e-05,
+      "loss": 3.9665,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334308418115481e-05,
+      "loss": 3.9852,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333469823364429e-05,
+      "loss": 3.9746,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33263286649375e-05,
+      "loss": 3.9807,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331795909623072e-05,
+      "loss": 3.9817,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33095731487202e-05,
+      "loss": 3.9648,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330118720120968e-05,
+      "loss": 3.9766,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329280125369915e-05,
+      "loss": 3.9626,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.328441530618863e-05,
+      "loss": 3.9955,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.327602935867811e-05,
+      "loss": 3.9636,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326764341116759e-05,
+      "loss": 3.987,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325925746365707e-05,
+      "loss": 3.9738,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325087151614655e-05,
+      "loss": 3.9706,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324250194743977e-05,
+      "loss": 3.9766,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323411599992925e-05,
+      "loss": 3.9767,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322573005241873e-05,
+      "loss": 3.9646,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3217360483711936e-05,
+      "loss": 3.9574,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3208974536201416e-05,
+      "loss": 3.9887,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3200588588690896e-05,
+      "loss": 3.9749,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3192202641180376e-05,
+      "loss": 3.9743,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3183816693669856e-05,
+      "loss": 3.9617,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3175430746159336e-05,
+      "loss": 3.9691,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3167061177452545e-05,
+      "loss": 3.9762,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3158675229942025e-05,
+      "loss": 3.9811,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3150289282431505e-05,
+      "loss": 3.9768,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3141903334920985e-05,
+      "loss": 3.9801,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3133517387410465e-05,
+      "loss": 3.9815,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3125131439899945e-05,
+      "loss": 3.9707,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.311674549238943e-05,
+      "loss": 3.9734,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310835954487891e-05,
+      "loss": 3.9718,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309998997617212e-05,
+      "loss": 3.9742,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30916040286616e-05,
+      "loss": 3.9734,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308321808115108e-05,
+      "loss": 3.9661,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307483213364056e-05,
+      "loss": 3.9822,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306644618613004e-05,
+      "loss": 3.9846,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3058060238619513e-05,
+      "loss": 3.973,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304967429110899e-05,
+      "loss": 3.9551,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304128834359847e-05,
+      "loss": 3.9763,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303291877489169e-05,
+      "loss": 3.9771,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302453282738117e-05,
+      "loss": 3.9797,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301614687987065e-05,
+      "loss": 3.9715,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300776093236013e-05,
+      "loss": 3.9748,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299937498484961e-05,
+      "loss": 3.9789,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299098903733909e-05,
+      "loss": 3.9658,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29826194686323e-05,
+      "loss": 3.9697,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297423352112178e-05,
+      "loss": 3.9756,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296584757361126e-05,
+      "loss": 3.9732,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295746162610074e-05,
+      "loss": 3.9842,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294910843619768e-05,
+      "loss": 3.9745,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294072248868716e-05,
+      "loss": 3.9646,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2932336541176636e-05,
+      "loss": 3.9677,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292395059366612e-05,
+      "loss": 3.9764,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29155646461556e-05,
+      "loss": 3.9788,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.290717869864508e-05,
+      "loss": 3.9668,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289879275113456e-05,
+      "loss": 3.9669,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289040680362404e-05,
+      "loss": 3.9616,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.288205361372099e-05,
+      "loss": 3.9774,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287366766621046e-05,
+      "loss": 3.9524,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286528171869994e-05,
+      "loss": 3.965,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285689577118942e-05,
+      "loss": 3.9694,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28485098236789e-05,
+      "loss": 3.9754,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284012387616838e-05,
+      "loss": 3.9633,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283173792865786e-05,
+      "loss": 3.9707,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282335198114734e-05,
+      "loss": 3.9614,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281496603363682e-05,
+      "loss": 3.9757,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2806612843733766e-05,
+      "loss": 3.9711,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2798226896223246e-05,
+      "loss": 3.9664,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2789840948712725e-05,
+      "loss": 3.972,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2781455001202205e-05,
+      "loss": 3.9848,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2773069053691685e-05,
+      "loss": 3.9677,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2764699484984894e-05,
+      "loss": 3.9691,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2756313537474374e-05,
+      "loss": 3.9708,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2747927589963854e-05,
+      "loss": 3.9708,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2739541642453334e-05,
+      "loss": 3.9708,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2731155694942814e-05,
+      "loss": 3.9709,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2722769747432294e-05,
+      "loss": 3.9675,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2714383799921774e-05,
+      "loss": 3.9734,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270599785241126e-05,
+      "loss": 3.9627,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269762828370447e-05,
+      "loss": 3.9646,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268924233619395e-05,
+      "loss": 3.9611,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268087276748716e-05,
+      "loss": 3.9663,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267248681997664e-05,
+      "loss": 3.9623,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266410087246612e-05,
+      "loss": 3.9802,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26557149249556e-05,
+      "loss": 3.981,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264734535624881e-05,
+      "loss": 3.9614,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263895940873829e-05,
+      "loss": 3.9737,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263057346122777e-05,
+      "loss": 3.9734,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262218751371725e-05,
+      "loss": 3.9753,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2613817945010464e-05,
+      "loss": 3.9632,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2605431997499944e-05,
+      "loss": 3.9655,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2597046049989424e-05,
+      "loss": 3.9759,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2588660102478904e-05,
+      "loss": 3.9768,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2580274154968384e-05,
+      "loss": 3.9618,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2571888207457864e-05,
+      "loss": 3.9712,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2563502259947344e-05,
+      "loss": 3.9775,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255513269124055e-05,
+      "loss": 3.9682,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254674674373003e-05,
+      "loss": 3.9642,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253836079621951e-05,
+      "loss": 3.9743,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252997484870899e-05,
+      "loss": 3.9657,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2521588901198466e-05,
+      "loss": 3.9741,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2513202953687946e-05,
+      "loss": 3.9751,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.046987533569336,
+      "eval_runtime": 307.3122,
+      "eval_samples_per_second": 1241.705,
+      "eval_steps_per_second": 38.804,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.250481700617743e-05,
+      "loss": 3.9665,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249644743747065e-05,
+      "loss": 3.955,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248806148996012e-05,
+      "loss": 3.9655,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24796755424496e-05,
+      "loss": 3.9799,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247128959493908e-05,
+      "loss": 3.9722,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246290364742856e-05,
+      "loss": 3.9718,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.245451769991804e-05,
+      "loss": 3.9664,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.244613175240752e-05,
+      "loss": 3.9694,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2437745804897e-05,
+      "loss": 3.9607,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242937623619021e-05,
+      "loss": 3.9701,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242099028867969e-05,
+      "loss": 3.965,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24126207199729e-05,
+      "loss": 3.9751,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2404251151266115e-05,
+      "loss": 3.9784,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2395865203755595e-05,
+      "loss": 3.9581,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2387479256245075e-05,
+      "loss": 3.9612,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2379093308734555e-05,
+      "loss": 3.9479,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2370707361224035e-05,
+      "loss": 3.9683,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2362321413713515e-05,
+      "loss": 3.9696,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2353951845006724e-05,
+      "loss": 3.9636,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2345565897496204e-05,
+      "loss": 3.9624,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2337179949985684e-05,
+      "loss": 3.9873,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2328794002475164e-05,
+      "loss": 3.9763,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2320408054964644e-05,
+      "loss": 3.9692,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2312022107454124e-05,
+      "loss": 3.9678,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2303636159943604e-05,
+      "loss": 3.9817,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2295250212433084e-05,
+      "loss": 3.9649,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.22868806437263e-05,
+      "loss": 3.9634,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227849469621578e-05,
+      "loss": 3.975,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227010874870526e-05,
+      "loss": 3.9606,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226172280119474e-05,
+      "loss": 3.9641,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.225333685368422e-05,
+      "loss": 3.9646,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.22449509061737e-05,
+      "loss": 3.9626,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.223656495866318e-05,
+      "loss": 3.972,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222817901115265e-05,
+      "loss": 3.977,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221980944244587e-05,
+      "loss": 3.9639,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221143987373908e-05,
+      "loss": 3.9736,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.220305392622856e-05,
+      "loss": 3.9727,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.219466797871804e-05,
+      "loss": 3.949,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2186282031207524e-05,
+      "loss": 3.9605,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2177896083697004e-05,
+      "loss": 3.966,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2169510136186484e-05,
+      "loss": 3.9509,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216112418867596e-05,
+      "loss": 3.9691,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.215273824116544e-05,
+      "loss": 3.9711,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.214435229365492e-05,
+      "loss": 3.9624,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.21359663461444e-05,
+      "loss": 3.9651,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2127596777437606e-05,
+      "loss": 3.9652,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211922720873082e-05,
+      "loss": 3.9626,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.21108412612203e-05,
+      "loss": 3.9748,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2102455313709775e-05,
+      "loss": 3.952,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209406936619926e-05,
+      "loss": 3.9693,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.208568341868874e-05,
+      "loss": 3.968,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207729747117822e-05,
+      "loss": 3.9642,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20689115236677e-05,
+      "loss": 3.9675,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206052557615718e-05,
+      "loss": 3.9518,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205215600745039e-05,
+      "loss": 3.9644,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.204377005993987e-05,
+      "loss": 3.9466,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203538411242935e-05,
+      "loss": 3.9808,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202699816491883e-05,
+      "loss": 3.9503,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201862859621204e-05,
+      "loss": 3.97,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201024264870152e-05,
+      "loss": 3.964,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2001856701191e-05,
+      "loss": 3.9592,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.199347075368048e-05,
+      "loss": 3.96,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1985101184973696e-05,
+      "loss": 3.9648,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1976715237463175e-05,
+      "loss": 3.9461,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1968329289952655e-05,
+      "loss": 3.9449,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1959943342442135e-05,
+      "loss": 3.9742,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1951573773735344e-05,
+      "loss": 3.9615,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1943187826224824e-05,
+      "loss": 3.9577,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1934801878714304e-05,
+      "loss": 3.947,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1926432310007514e-05,
+      "loss": 3.9575,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1918046362496993e-05,
+      "loss": 3.9632,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1909660414986473e-05,
+      "loss": 3.9661,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190127446747595e-05,
+      "loss": 3.9667,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189288851996543e-05,
+      "loss": 3.9661,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188451895125865e-05,
+      "loss": 3.966,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187613300374813e-05,
+      "loss": 3.9599,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186774705623761e-05,
+      "loss": 3.9583,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185936110872709e-05,
+      "loss": 3.9572,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185097516121657e-05,
+      "loss": 3.9641,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.184258921370605e-05,
+      "loss": 3.9597,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.183420326619553e-05,
+      "loss": 3.9527,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182585007629247e-05,
+      "loss": 3.9669,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.181746412878195e-05,
+      "loss": 3.969,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.180907818127143e-05,
+      "loss": 3.9605,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.180069223376091e-05,
+      "loss": 3.9493,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.179230628625039e-05,
+      "loss": 3.9562,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.178392033873987e-05,
+      "loss": 3.963,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.177553439122935e-05,
+      "loss": 3.9695,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1767148443718834e-05,
+      "loss": 3.9592,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.175877887501204e-05,
+      "loss": 3.9612,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.175039292750152e-05,
+      "loss": 3.9637,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1742006979991e-05,
+      "loss": 3.9573,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173362103248048e-05,
+      "loss": 3.9528,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172523508496996e-05,
+      "loss": 3.9592,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1716849137459436e-05,
+      "loss": 3.9634,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1708463189948916e-05,
+      "loss": 3.9667,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1700077242438396e-05,
+      "loss": 3.9656,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1691691294927876e-05,
+      "loss": 3.9498,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1683305347417356e-05,
+      "loss": 3.9568,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.167493577871057e-05,
+      "loss": 3.9596,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.166654983120005e-05,
+      "loss": 3.9656,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.165816388368953e-05,
+      "loss": 3.9561,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164977793617901e-05,
+      "loss": 3.9553,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164139198866849e-05,
+      "loss": 3.9497,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.163300604115797e-05,
+      "loss": 3.9584,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.162462009364745e-05,
+      "loss": 3.9411,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161623414613693e-05,
+      "loss": 3.9555,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160786457743014e-05,
+      "loss": 3.958,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159947862991962e-05,
+      "loss": 3.9591,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15910926824091e-05,
+      "loss": 3.9513,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.158270673489858e-05,
+      "loss": 3.9581,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157432078738806e-05,
+      "loss": 3.9504,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156595121868127e-05,
+      "loss": 3.9596,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1557565271170756e-05,
+      "loss": 3.9615,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1549179323660236e-05,
+      "loss": 3.9522,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1540793376149716e-05,
+      "loss": 3.9615,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1532423807442925e-05,
+      "loss": 3.9685,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1524054238736134e-05,
+      "loss": 3.9627,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1515668291225614e-05,
+      "loss": 3.948,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1507282343715094e-05,
+      "loss": 3.9631,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1498896396204574e-05,
+      "loss": 3.9596,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1490510448694054e-05,
+      "loss": 3.9522,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1482124501183534e-05,
+      "loss": 3.9603,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1473738553673014e-05,
+      "loss": 3.9542,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1465352606162494e-05,
+      "loss": 3.9596,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1456966658651974e-05,
+      "loss": 3.9522,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144859708994519e-05,
+      "loss": 3.9551,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144021114243467e-05,
+      "loss": 3.9456,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.143182519492415e-05,
+      "loss": 3.9535,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.142343924741362e-05,
+      "loss": 3.9436,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14150532999031e-05,
+      "loss": 3.9667,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140666735239258e-05,
+      "loss": 3.9692,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139828140488206e-05,
+      "loss": 3.9526,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138989545737154e-05,
+      "loss": 3.9556,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138152588866475e-05,
+      "loss": 3.9634,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.137313994115423e-05,
+      "loss": 3.9606,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.136475399364371e-05,
+      "loss": 3.9532,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135636804613319e-05,
+      "loss": 3.9514,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134799847742641e-05,
+      "loss": 3.9639,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133961252991589e-05,
+      "loss": 3.964,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133122658240537e-05,
+      "loss": 3.952,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.132284063489485e-05,
+      "loss": 3.9575,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1314471066188056e-05,
+      "loss": 3.9643,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1306085118677536e-05,
+      "loss": 3.9524,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1297699171167016e-05,
+      "loss": 3.9532,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1289313223656496e-05,
+      "loss": 3.9627,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1280927276145976e-05,
+      "loss": 3.9536,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1272541328635456e-05,
+      "loss": 3.9609,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1264188138732394e-05,
+      "loss": 3.9596,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0399980545043945,
+      "eval_runtime": 307.7992,
+      "eval_samples_per_second": 1239.74,
+      "eval_steps_per_second": 38.743,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125580219122188e-05,
+      "loss": 3.9349,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124741624371136e-05,
+      "loss": 3.9456,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123903029620084e-05,
+      "loss": 3.9532,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123064434869032e-05,
+      "loss": 3.9666,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.12222584011798e-05,
+      "loss": 3.9584,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121387245366928e-05,
+      "loss": 3.9658,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120548650615876e-05,
+      "loss": 3.9495,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119710055864824e-05,
+      "loss": 3.9599,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118871461113772e-05,
+      "loss": 3.945,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11803286636272e-05,
+      "loss": 3.9581,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117194271611668e-05,
+      "loss": 3.9519,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116355676860616e-05,
+      "loss": 3.964,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.115518719989937e-05,
+      "loss": 3.9664,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114680125238885e-05,
+      "loss": 3.9456,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113841530487833e-05,
+      "loss": 3.9469,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113002935736781e-05,
+      "loss": 3.9407,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112164340985729e-05,
+      "loss": 3.9501,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111325746234677e-05,
+      "loss": 3.9572,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110487151483625e-05,
+      "loss": 3.9548,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109648556732573e-05,
+      "loss": 3.9491,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108809961981521e-05,
+      "loss": 3.9782,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107971367230469e-05,
+      "loss": 3.9619,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107132772479417e-05,
+      "loss": 3.9612,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.106294177728365e-05,
+      "loss": 3.955,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.105455582977313e-05,
+      "loss": 3.9705,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.104616988226261e-05,
+      "loss": 3.9529,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103780031355582e-05,
+      "loss": 3.9552,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.10294143660453e-05,
+      "loss": 3.9624,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102102841853478e-05,
+      "loss": 3.9467,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1012642471024265e-05,
+      "loss": 3.9563,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1004256523513745e-05,
+      "loss": 3.9507,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0995886954806954e-05,
+      "loss": 3.9533,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0987501007296434e-05,
+      "loss": 3.9589,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0979115059785914e-05,
+      "loss": 3.9646,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0970729112275394e-05,
+      "loss": 3.9545,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09623595435686e-05,
+      "loss": 3.9647,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.095398997486181e-05,
+      "loss": 3.9602,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094560402735129e-05,
+      "loss": 3.9352,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093721807984077e-05,
+      "loss": 3.9521,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092883213233025e-05,
+      "loss": 3.951,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092044618481973e-05,
+      "loss": 3.9373,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091206023730921e-05,
+      "loss": 3.9587,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09036742897987e-05,
+      "loss": 3.9584,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089528834228817e-05,
+      "loss": 3.9552,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088691877358139e-05,
+      "loss": 3.9466,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087853282607087e-05,
+      "loss": 3.9579,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087014687856035e-05,
+      "loss": 3.9507,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086176093104982e-05,
+      "loss": 3.9671,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0853391362343036e-05,
+      "loss": 3.9391,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0845005414832516e-05,
+      "loss": 3.9548,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0836619467321996e-05,
+      "loss": 3.9596,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082823351981147e-05,
+      "loss": 3.9497,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081984757230095e-05,
+      "loss": 3.9598,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0811478003594165e-05,
+      "loss": 3.9376,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0803092056083645e-05,
+      "loss": 3.9541,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0794706108573125e-05,
+      "loss": 3.9367,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0786320161062605e-05,
+      "loss": 3.9668,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077795059235582e-05,
+      "loss": 3.9424,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0769564644845294e-05,
+      "loss": 3.957,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0761178697334774e-05,
+      "loss": 3.9539,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0752792749824254e-05,
+      "loss": 3.946,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0744406802313734e-05,
+      "loss": 3.9481,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0736020854803214e-05,
+      "loss": 3.9539,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072765128609642e-05,
+      "loss": 3.9346,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071928171738964e-05,
+      "loss": 3.9348,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071089576987912e-05,
+      "loss": 3.958,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07025098223686e-05,
+      "loss": 3.951,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069412387485808e-05,
+      "loss": 3.9498,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.068573792734756e-05,
+      "loss": 3.9368,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067735197983704e-05,
+      "loss": 3.943,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066896603232652e-05,
+      "loss": 3.9506,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0660580084816e-05,
+      "loss": 3.9537,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.065219413730548e-05,
+      "loss": 3.9563,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.064382456859869e-05,
+      "loss": 3.9519,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.063543862108817e-05,
+      "loss": 3.9578,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.062705267357765e-05,
+      "loss": 3.9497,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061866672606713e-05,
+      "loss": 3.9457,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0610297157360343e-05,
+      "loss": 3.9485,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0601911209849823e-05,
+      "loss": 3.9481,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05935252623393e-05,
+      "loss": 3.951,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058513931482878e-05,
+      "loss": 3.9463,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057675336731826e-05,
+      "loss": 3.9495,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056838379861147e-05,
+      "loss": 3.9638,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055999785110095e-05,
+      "loss": 3.9471,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055161190359043e-05,
+      "loss": 3.9381,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054322595607991e-05,
+      "loss": 3.945,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053485638737312e-05,
+      "loss": 3.9527,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05264704398626e-05,
+      "loss": 3.9563,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051810087115581e-05,
+      "loss": 3.9488,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509731302449023e-05,
+      "loss": 3.9476,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0501345354938503e-05,
+      "loss": 3.9561,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0492959407427986e-05,
+      "loss": 3.9457,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0484573459917466e-05,
+      "loss": 3.9415,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0476187512406946e-05,
+      "loss": 3.9467,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0467801564896426e-05,
+      "loss": 3.95,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0459415617385906e-05,
+      "loss": 3.9562,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0451029669875386e-05,
+      "loss": 3.9534,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442660101168595e-05,
+      "loss": 3.9404,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0434274153658075e-05,
+      "loss": 3.9443,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.042588820614756e-05,
+      "loss": 3.9471,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041750225863704e-05,
+      "loss": 3.9576,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0409116311126518e-05,
+      "loss": 3.9424,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0400730363615998e-05,
+      "loss": 3.9429,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0392344416105478e-05,
+      "loss": 3.9423,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.038395846859495e-05,
+      "loss": 3.9471,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375588899888167e-05,
+      "loss": 3.9306,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036720295237765e-05,
+      "loss": 3.9444,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035881700486713e-05,
+      "loss": 3.9421,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0350431057356604e-05,
+      "loss": 3.9481,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.034206148864982e-05,
+      "loss": 3.9403,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.03336755411393e-05,
+      "loss": 3.9487,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0325289593628776e-05,
+      "loss": 3.9363,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0316903646118256e-05,
+      "loss": 3.9484,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0308534077411472e-05,
+      "loss": 3.9486,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0300148129900952e-05,
+      "loss": 3.9385,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0291762182390425e-05,
+      "loss": 3.9514,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0283376234879905e-05,
+      "loss": 3.9569,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.027500666617312e-05,
+      "loss": 3.9512,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0266637097466333e-05,
+      "loss": 3.9414,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0258251149955813e-05,
+      "loss": 3.9448,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0249865202445293e-05,
+      "loss": 3.9501,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0241479254934773e-05,
+      "loss": 3.9402,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.023309330742425e-05,
+      "loss": 3.9541,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.022470735991373e-05,
+      "loss": 3.9423,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0216337791206946e-05,
+      "loss": 3.9527,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0207951843696426e-05,
+      "loss": 3.9385,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.01995658961859e-05,
+      "loss": 3.9448,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.019117994867538e-05,
+      "loss": 3.9356,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182810379968595e-05,
+      "loss": 3.941,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174424432458075e-05,
+      "loss": 3.9325,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.016603848494755e-05,
+      "loss": 3.9535,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.015765253743703e-05,
+      "loss": 3.9627,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014926658992651e-05,
+      "loss": 3.9434,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014088064241599e-05,
+      "loss": 3.9405,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.013249469490547e-05,
+      "loss": 3.953,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.012410874739495e-05,
+      "loss": 3.9513,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0115739178688163e-05,
+      "loss": 3.9464,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0107353231177643e-05,
+      "loss": 3.9386,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0098967283667123e-05,
+      "loss": 3.9555,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0090581336156603e-05,
+      "loss": 3.9478,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0082211767449812e-05,
+      "loss": 3.9443,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0073825819939296e-05,
+      "loss": 3.9426,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0065439872428776e-05,
+      "loss": 3.9515,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0057070303721985e-05,
+      "loss": 3.9426,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0048684356211465e-05,
+      "loss": 3.9418,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0040298408700945e-05,
+      "loss": 3.9516,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0031912461190425e-05,
+      "loss": 3.9429,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023526513679905e-05,
+      "loss": 3.9455,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0015156944973117e-05,
+      "loss": 3.9515,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0006770997462597e-05,
+      "loss": 3.9436,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0352654457092285,
+      "eval_runtime": 297.5559,
+      "eval_samples_per_second": 1282.418,
+      "eval_steps_per_second": 40.077,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998385049952077e-05,
+      "loss": 3.9348,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989999102441557e-05,
+      "loss": 3.9403,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981613154931037e-05,
+      "loss": 3.9546,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9973227207420517e-05,
+      "loss": 3.95,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964841259909997e-05,
+      "loss": 3.9544,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956455312399477e-05,
+      "loss": 3.9388,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.994806936488896e-05,
+      "loss": 3.9513,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.993968341737844e-05,
+      "loss": 3.9307,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931297469867913e-05,
+      "loss": 3.9537,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922911522357393e-05,
+      "loss": 3.9363,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914525574846873e-05,
+      "loss": 3.956,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9906139627336353e-05,
+      "loss": 3.9543,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9897770058629566e-05,
+      "loss": 3.9383,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889384111119046e-05,
+      "loss": 3.9396,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880998163608525e-05,
+      "loss": 3.9321,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9872612216098005e-05,
+      "loss": 3.9335,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9864226268587485e-05,
+      "loss": 3.9475,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9855840321076965e-05,
+      "loss": 3.9415,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9847454373566445e-05,
+      "loss": 3.943,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9839084804859658e-05,
+      "loss": 3.9655,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830698857349138e-05,
+      "loss": 3.9491,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822312909838618e-05,
+      "loss": 3.9534,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813926962328098e-05,
+      "loss": 3.9423,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9805541014817578e-05,
+      "loss": 3.9576,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9797155067307058e-05,
+      "loss": 3.9404,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788769119796538e-05,
+      "loss": 3.9506,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978038317228602e-05,
+      "loss": 3.9484,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.97719972247755e-05,
+      "loss": 3.9406,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763611277264974e-05,
+      "loss": 3.9412,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755225329754454e-05,
+      "loss": 3.9394,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.974685576104767e-05,
+      "loss": 3.9444,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973846981353715e-05,
+      "loss": 3.9464,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9730083866026626e-05,
+      "loss": 3.9522,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9721697918516106e-05,
+      "loss": 3.9472,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9713328349809322e-05,
+      "loss": 3.9503,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9704942402298795e-05,
+      "loss": 3.9481,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9696556454788275e-05,
+      "loss": 3.9293,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968817050727776e-05,
+      "loss": 3.9414,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967978455976724e-05,
+      "loss": 3.9369,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967139861225672e-05,
+      "loss": 3.9296,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96630126647462e-05,
+      "loss": 3.9456,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9654643096039408e-05,
+      "loss": 3.9498,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9646257148528888e-05,
+      "loss": 3.9398,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9637871201018368e-05,
+      "loss": 3.9398,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962948525350785e-05,
+      "loss": 3.9472,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962109930599733e-05,
+      "loss": 3.9449,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961271335848681e-05,
+      "loss": 3.9497,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.960432741097629e-05,
+      "loss": 3.9346,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.959594146346577e-05,
+      "loss": 3.9395,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.958755551595525e-05,
+      "loss": 3.9461,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957918594724846e-05,
+      "loss": 3.9416,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9570816378541672e-05,
+      "loss": 3.9436,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562430431031152e-05,
+      "loss": 3.9314,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9554044483520632e-05,
+      "loss": 3.9448,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545658536010112e-05,
+      "loss": 3.9248,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.953728896730332e-05,
+      "loss": 3.952,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528903019792805e-05,
+      "loss": 3.9361,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520517072282285e-05,
+      "loss": 3.943,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9512131124771764e-05,
+      "loss": 3.9401,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503745177261244e-05,
+      "loss": 3.9397,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9495375608554454e-05,
+      "loss": 3.9333,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486989661043933e-05,
+      "loss": 3.9444,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478603713533413e-05,
+      "loss": 3.9265,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9470217766022897e-05,
+      "loss": 3.9243,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9461831818512377e-05,
+      "loss": 3.945,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9453462249805586e-05,
+      "loss": 3.946,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9445076302295066e-05,
+      "loss": 3.9371,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9436690354784546e-05,
+      "loss": 3.9235,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9428304407274026e-05,
+      "loss": 3.9326,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9419918459763506e-05,
+      "loss": 3.9414,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411532512252986e-05,
+      "loss": 3.9465,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9403146564742462e-05,
+      "loss": 3.942,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394760617231942e-05,
+      "loss": 3.9422,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386391048525158e-05,
+      "loss": 3.9467,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9378021479818367e-05,
+      "loss": 3.9446,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369635532307847e-05,
+      "loss": 3.9298,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.936124958479733e-05,
+      "loss": 3.9359,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.935286363728681e-05,
+      "loss": 3.9419,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344477689776284e-05,
+      "loss": 3.9364,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9336091742265763e-05,
+      "loss": 3.9359,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9327705794755243e-05,
+      "loss": 3.9372,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931933622604846e-05,
+      "loss": 3.9553,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310950278537936e-05,
+      "loss": 3.9352,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302580709831152e-05,
+      "loss": 3.9255,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9294194762320632e-05,
+      "loss": 3.9365,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9285808814810105e-05,
+      "loss": 3.9401,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9277422867299585e-05,
+      "loss": 3.9459,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9269036919789068e-05,
+      "loss": 3.935,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260650972278548e-05,
+      "loss": 3.9402,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252265024768028e-05,
+      "loss": 3.9412,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243879077257508e-05,
+      "loss": 3.9402,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235493129746988e-05,
+      "loss": 3.9283,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9227107182236468e-05,
+      "loss": 3.9357,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9218737613529677e-05,
+      "loss": 3.9449,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921035166601916e-05,
+      "loss": 3.9492,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.920196571850864e-05,
+      "loss": 3.9357,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.919357977099812e-05,
+      "loss": 3.9304,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91851938234876e-05,
+      "loss": 3.9365,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.917680787597708e-05,
+      "loss": 3.9339,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916842192846656e-05,
+      "loss": 3.9459,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916003598095604e-05,
+      "loss": 3.9329,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915165003344552e-05,
+      "loss": 3.9332,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143280464738733e-05,
+      "loss": 3.9288,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9134894517228213e-05,
+      "loss": 3.9392,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126508569717692e-05,
+      "loss": 3.9157,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9118122622207172e-05,
+      "loss": 3.9342,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910975305350038e-05,
+      "loss": 3.9297,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910136710598986e-05,
+      "loss": 3.9397,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092981158479345e-05,
+      "loss": 3.9308,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9084595210968818e-05,
+      "loss": 3.9385,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076209263458298e-05,
+      "loss": 3.9285,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067839694751514e-05,
+      "loss": 3.9349,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059453747240994e-05,
+      "loss": 3.9365,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905106779973047e-05,
+      "loss": 3.929,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904268185221995e-05,
+      "loss": 3.9439,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9034312283513166e-05,
+      "loss": 3.9488,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025942714806375e-05,
+      "loss": 3.94,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9017556767295855e-05,
+      "loss": 3.9306,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9009170819785335e-05,
+      "loss": 3.9335,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000784872274815e-05,
+      "loss": 3.9378,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992398924764292e-05,
+      "loss": 3.9331,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898401297725377e-05,
+      "loss": 3.939,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897562702974325e-05,
+      "loss": 3.9325,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.896724108223273e-05,
+      "loss": 3.94,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.895885513472221e-05,
+      "loss": 3.9289,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950485566015424e-05,
+      "loss": 3.9337,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8942099618504904e-05,
+      "loss": 3.9226,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933713670994384e-05,
+      "loss": 3.9285,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925327723483864e-05,
+      "loss": 3.9227,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916958154777073e-05,
+      "loss": 3.9452,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908572207266553e-05,
+      "loss": 3.9515,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8900186259756036e-05,
+      "loss": 3.9408,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891800312245516e-05,
+      "loss": 3.9225,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883430743538725e-05,
+      "loss": 3.9432,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8875044796028205e-05,
+      "loss": 3.9398,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8866658848517685e-05,
+      "loss": 3.9345,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8858272901007165e-05,
+      "loss": 3.9293,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849886953496645e-05,
+      "loss": 3.943,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841517384789858e-05,
+      "loss": 3.9373,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8833131437279338e-05,
+      "loss": 3.9328,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8824745489768818e-05,
+      "loss": 3.9372,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8816375921062027e-05,
+      "loss": 3.9395,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807989973551507e-05,
+      "loss": 3.93,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879960402604099e-05,
+      "loss": 3.9324,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879121807853047e-05,
+      "loss": 3.9429,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.878283213101995e-05,
+      "loss": 3.9326,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.877444618350943e-05,
+      "loss": 3.9348,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.876607661480264e-05,
+      "loss": 3.9443,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.875769066729212e-05,
+      "loss": 3.9289,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.029688358306885,
+      "eval_runtime": 293.1449,
+      "eval_samples_per_second": 1301.714,
+      "eval_steps_per_second": 40.68,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.87493047197816e-05,
+      "loss": 3.9231,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.874091877227108e-05,
+      "loss": 3.9307,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8732532824760562e-05,
+      "loss": 3.9447,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8724146877250042e-05,
+      "loss": 3.9367,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8715760929739522e-05,
+      "loss": 3.9461,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8707374982229002e-05,
+      "loss": 3.9283,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8698989034718482e-05,
+      "loss": 3.9385,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8690603087207955e-05,
+      "loss": 3.9211,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.868221713969744e-05,
+      "loss": 3.9425,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.867383119218692e-05,
+      "loss": 3.9263,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.86654452446764e-05,
+      "loss": 3.9444,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.865705929716588e-05,
+      "loss": 3.9392,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8648689728459087e-05,
+      "loss": 3.9336,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8640303780948567e-05,
+      "loss": 3.9259,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8631917833438047e-05,
+      "loss": 3.9228,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.862353188592753e-05,
+      "loss": 3.9243,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.861514593841701e-05,
+      "loss": 3.9329,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.860675999090649e-05,
+      "loss": 3.9324,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.859837404339597e-05,
+      "loss": 3.9286,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858998809588545e-05,
+      "loss": 3.9527,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858160214837493e-05,
+      "loss": 3.9419,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.857321620086441e-05,
+      "loss": 3.9376,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8564846632157623e-05,
+      "loss": 3.9319,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8556460684647103e-05,
+      "loss": 3.9497,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8548074737136583e-05,
+      "loss": 3.9256,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8539688789626063e-05,
+      "loss": 3.9386,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8531302842115543e-05,
+      "loss": 3.9369,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8522933273408752e-05,
+      "loss": 3.9349,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8514547325898232e-05,
+      "loss": 3.9264,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8506177757191444e-05,
+      "loss": 3.9297,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8497791809680924e-05,
+      "loss": 3.9348,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8489405862170404e-05,
+      "loss": 3.9343,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8481019914659884e-05,
+      "loss": 3.9385,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8472633967149364e-05,
+      "loss": 3.9386,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8464264398442577e-05,
+      "loss": 3.9393,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8455878450932057e-05,
+      "loss": 3.9344,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8447492503421537e-05,
+      "loss": 3.9221,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8439106555911016e-05,
+      "loss": 3.933,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843072060840049e-05,
+      "loss": 3.9258,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.842233466088997e-05,
+      "loss": 3.9213,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8413948713379453e-05,
+      "loss": 3.936,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8405562765868933e-05,
+      "loss": 3.9402,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8397176818358413e-05,
+      "loss": 3.9267,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8388790870847893e-05,
+      "loss": 3.9289,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8380404923337373e-05,
+      "loss": 3.9334,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8372018975826853e-05,
+      "loss": 3.9353,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8363649407120062e-05,
+      "loss": 3.9387,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8355279838413274e-05,
+      "loss": 3.9285,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8346893890902754e-05,
+      "loss": 3.9301,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8338507943392234e-05,
+      "loss": 3.9345,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8330121995881714e-05,
+      "loss": 3.9283,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8321736048371194e-05,
+      "loss": 3.9321,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8313350100860674e-05,
+      "loss": 3.9272,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8304964153350154e-05,
+      "loss": 3.9352,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8296578205839634e-05,
+      "loss": 3.9127,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8288208637132846e-05,
+      "loss": 3.9377,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8279822689622326e-05,
+      "loss": 3.9274,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8271436742111806e-05,
+      "loss": 3.9305,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8263050794601286e-05,
+      "loss": 3.9358,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8254681225894495e-05,
+      "loss": 3.9285,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.824629527838398e-05,
+      "loss": 3.9223,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.823790933087346e-05,
+      "loss": 3.9383,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8229539762166668e-05,
+      "loss": 3.9121,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8221153814656148e-05,
+      "loss": 3.9148,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8212767867145628e-05,
+      "loss": 3.9324,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8204381919635108e-05,
+      "loss": 3.9353,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8195995972124588e-05,
+      "loss": 3.9282,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.818761002461407e-05,
+      "loss": 3.9156,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.817922407710355e-05,
+      "loss": 3.9201,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.817085450839676e-05,
+      "loss": 3.9264,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.816246856088624e-05,
+      "loss": 3.9357,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.815408261337572e-05,
+      "loss": 3.9295,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.81456966658652e-05,
+      "loss": 3.9326,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8137310718354676e-05,
+      "loss": 3.9336,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8128924770844156e-05,
+      "loss": 3.9372,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8120538823333636e-05,
+      "loss": 3.9182,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8112152875823116e-05,
+      "loss": 3.924,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8103783307116325e-05,
+      "loss": 3.9353,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.809539735960581e-05,
+      "loss": 3.923,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8087027790899025e-05,
+      "loss": 3.9286,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8078641843388498e-05,
+      "loss": 3.9217,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8070255895877978e-05,
+      "loss": 3.9458,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8061869948367458e-05,
+      "loss": 3.9254,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8053484000856938e-05,
+      "loss": 3.9153,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.804511443215015e-05,
+      "loss": 3.9264,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.803672848463963e-05,
+      "loss": 3.9297,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.802834253712911e-05,
+      "loss": 3.9356,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.801995658961859e-05,
+      "loss": 3.9255,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.801157064210807e-05,
+      "loss": 3.9275,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.800318469459755e-05,
+      "loss": 3.9334,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.799479874708703e-05,
+      "loss": 3.9258,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.798641279957651e-05,
+      "loss": 3.9208,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7978026852065993e-05,
+      "loss": 3.9254,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7969657283359202e-05,
+      "loss": 3.9353,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7961271335848682e-05,
+      "loss": 3.9339,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7952885388338162e-05,
+      "loss": 3.9317,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7944499440827642e-05,
+      "loss": 3.9149,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7936129872120855e-05,
+      "loss": 3.927,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7927743924610335e-05,
+      "loss": 3.9259,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7919357977099815e-05,
+      "loss": 3.9315,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7910988408393024e-05,
+      "loss": 3.9255,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7902602460882504e-05,
+      "loss": 3.9225,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7894216513371984e-05,
+      "loss": 3.9177,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7885830565861464e-05,
+      "loss": 3.9324,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7877444618350947e-05,
+      "loss": 3.9007,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7869058670840427e-05,
+      "loss": 3.9255,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7860672723329907e-05,
+      "loss": 3.9181,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7852286775819387e-05,
+      "loss": 3.9308,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7843917207112596e-05,
+      "loss": 3.9175,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.783554763840581e-05,
+      "loss": 3.9301,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7827161690895288e-05,
+      "loss": 3.9167,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7818775743384768e-05,
+      "loss": 3.9209,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7810389795874248e-05,
+      "loss": 3.9275,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7802003848363728e-05,
+      "loss": 3.9199,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7793617900853208e-05,
+      "loss": 3.932,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7785231953342688e-05,
+      "loss": 3.9368,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7776846005832165e-05,
+      "loss": 3.9281,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7768460058321645e-05,
+      "loss": 3.9202,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7760074110811125e-05,
+      "loss": 3.9261,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7751688163300605e-05,
+      "loss": 3.9296,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7743302215790084e-05,
+      "loss": 3.9155,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7734916268279564e-05,
+      "loss": 3.9356,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7726546699572774e-05,
+      "loss": 3.922,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7718177130865986e-05,
+      "loss": 3.9277,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7709791183355466e-05,
+      "loss": 3.9233,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7701405235844946e-05,
+      "loss": 3.9221,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7693019288334426e-05,
+      "loss": 3.9099,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7684633340823906e-05,
+      "loss": 3.9187,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7676247393313386e-05,
+      "loss": 3.912,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7667861445802866e-05,
+      "loss": 3.9337,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.765947549829235e-05,
+      "loss": 3.9445,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7651105929585558e-05,
+      "loss": 3.9289,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7642719982075038e-05,
+      "loss": 3.9125,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7634334034564518e-05,
+      "loss": 3.9326,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7625948087053998e-05,
+      "loss": 3.9244,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.761757851834721e-05,
+      "loss": 3.9262,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.760920894964042e-05,
+      "loss": 3.9225,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76008230021299e-05,
+      "loss": 3.9288,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.759243705461938e-05,
+      "loss": 3.9229,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.758405110710886e-05,
+      "loss": 3.9243,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.757566515959834e-05,
+      "loss": 3.9261,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.756727921208782e-05,
+      "loss": 3.9285,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7558893264577303e-05,
+      "loss": 3.9244,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7550523695870512e-05,
+      "loss": 3.9172,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7542137748359992e-05,
+      "loss": 3.9327,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7533751800849472e-05,
+      "loss": 3.9273,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7525365853338952e-05,
+      "loss": 3.9211,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7516996284632164e-05,
+      "loss": 3.9366,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7508610337121644e-05,
+      "loss": 3.9188,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.024346828460693,
+      "eval_runtime": 295.8513,
+      "eval_samples_per_second": 1289.807,
+      "eval_steps_per_second": 40.307,
+      "step": 1373760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7500224389611124e-05,
+      "loss": 3.9171,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7491838442100604e-05,
+      "loss": 3.9197,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7483452494590084e-05,
+      "loss": 3.9342,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7475066547079564e-05,
+      "loss": 3.927,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7466680599569044e-05,
+      "loss": 3.9345,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7458311030862256e-05,
+      "loss": 3.9197,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7449925083351736e-05,
+      "loss": 3.9278,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7441539135841216e-05,
+      "loss": 3.9089,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7433153188330696e-05,
+      "loss": 3.9338,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.742476724082017e-05,
+      "loss": 3.9139,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.741638129330965e-05,
+      "loss": 3.9323,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7407995345799133e-05,
+      "loss": 3.9281,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739962577709235e-05,
+      "loss": 3.9261,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7391239829581822e-05,
+      "loss": 3.9154,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7382853882071302e-05,
+      "loss": 3.9162,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7374467934560782e-05,
+      "loss": 3.9163,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.736608198705026e-05,
+      "loss": 3.9213,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7357712418343474e-05,
+      "loss": 3.9231,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7349326470832954e-05,
+      "loss": 3.9234,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7340940523322434e-05,
+      "loss": 3.9407,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7332554575811914e-05,
+      "loss": 3.9342,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7324168628301394e-05,
+      "loss": 3.9281,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7315782680790874e-05,
+      "loss": 3.924,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7307396733280354e-05,
+      "loss": 3.9343,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7299027164573566e-05,
+      "loss": 3.9198,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7290641217063046e-05,
+      "loss": 3.9288,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7282255269552526e-05,
+      "loss": 3.9255,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7273869322042006e-05,
+      "loss": 3.9264,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7265499753335215e-05,
+      "loss": 3.9162,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7257113805824695e-05,
+      "loss": 3.9208,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.724872785831418e-05,
+      "loss": 3.925,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.724034191080366e-05,
+      "loss": 3.9271,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.723195596329314e-05,
+      "loss": 3.9332,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.722357001578262e-05,
+      "loss": 3.928,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7215200447075828e-05,
+      "loss": 3.928,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7206814499565308e-05,
+      "loss": 3.9236,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7198428552054788e-05,
+      "loss": 3.913,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7190042604544267e-05,
+      "loss": 3.9256,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.718165665703375e-05,
+      "loss": 3.9156,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.717327070952323e-05,
+      "loss": 3.915,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.716488476201271e-05,
+      "loss": 3.9215,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7156498814502184e-05,
+      "loss": 3.9325,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7148112866991664e-05,
+      "loss": 3.9179,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.713974329828488e-05,
+      "loss": 3.9196,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7131373729578092e-05,
+      "loss": 3.9268,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7122987782067572e-05,
+      "loss": 3.9216,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7114601834557052e-05,
+      "loss": 3.9329,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7106215887046532e-05,
+      "loss": 3.9176,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7097829939536005e-05,
+      "loss": 3.9151,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.708946037082922e-05,
+      "loss": 3.9271,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7081074423318704e-05,
+      "loss": 3.9201,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7072688475808184e-05,
+      "loss": 3.9258,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7064302528297658e-05,
+      "loss": 3.9181,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7055916580787138e-05,
+      "loss": 3.9214,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7047530633276618e-05,
+      "loss": 3.9114,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.703916106456983e-05,
+      "loss": 3.9236,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.703077511705931e-05,
+      "loss": 3.9206,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.702238916954879e-05,
+      "loss": 3.9208,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.701400322203827e-05,
+      "loss": 3.9278,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.700561727452775e-05,
+      "loss": 3.9169,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.699723132701723e-05,
+      "loss": 3.9132,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698884537950671e-05,
+      "loss": 3.9333,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698045943199619e-05,
+      "loss": 3.9041,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6972089863289402e-05,
+      "loss": 3.9037,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6963703915778882e-05,
+      "loss": 3.9199,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6955317968268362e-05,
+      "loss": 3.9304,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6946932020757842e-05,
+      "loss": 3.9185,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6938546073247322e-05,
+      "loss": 3.9102,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6930176504540534e-05,
+      "loss": 3.912,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6921790557030014e-05,
+      "loss": 3.9139,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6913404609519494e-05,
+      "loss": 3.9284,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6905018662008974e-05,
+      "loss": 3.9198,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6896632714498454e-05,
+      "loss": 3.9238,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6888263145791663e-05,
+      "loss": 3.9241,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6879877198281143e-05,
+      "loss": 3.9306,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6871491250770627e-05,
+      "loss": 3.91,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6863121682063836e-05,
+      "loss": 3.9147,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6854735734553316e-05,
+      "loss": 3.928,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6846349787042796e-05,
+      "loss": 3.9122,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6837963839532276e-05,
+      "loss": 3.9208,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6829577892021756e-05,
+      "loss": 3.9127,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6821191944511236e-05,
+      "loss": 3.933,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.681280599700072e-05,
+      "loss": 3.9188,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6804436428293928e-05,
+      "loss": 3.9091,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6796050480783408e-05,
+      "loss": 3.9125,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6787664533272888e-05,
+      "loss": 3.9198,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6779278585762368e-05,
+      "loss": 3.9306,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6770892638251844e-05,
+      "loss": 3.9175,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.676252306954506e-05,
+      "loss": 3.9207,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.675413712203454e-05,
+      "loss": 3.9143,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6745751174524013e-05,
+      "loss": 3.9251,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6737365227013493e-05,
+      "loss": 3.9142,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6728979279502973e-05,
+      "loss": 3.9118,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6720593331992457e-05,
+      "loss": 3.9266,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6712223763285666e-05,
+      "loss": 3.9258,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6703837815775146e-05,
+      "loss": 3.9248,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6695451868264626e-05,
+      "loss": 3.9087,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6687065920754106e-05,
+      "loss": 3.9188,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6678679973243586e-05,
+      "loss": 3.9155,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6670294025733066e-05,
+      "loss": 3.9215,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6661908078222546e-05,
+      "loss": 3.9188,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6653538509515758e-05,
+      "loss": 3.912,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6645152562005238e-05,
+      "loss": 3.91,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6636766614494718e-05,
+      "loss": 3.923,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6628397045787927e-05,
+      "loss": 3.8971,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6620011098277407e-05,
+      "loss": 3.9145,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.661162515076689e-05,
+      "loss": 3.9064,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.660323920325637e-05,
+      "loss": 3.9227,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.659485325574585e-05,
+      "loss": 3.9125,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.658646730823533e-05,
+      "loss": 3.9186,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.657808136072481e-05,
+      "loss": 3.9102,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.656969541321429e-05,
+      "loss": 3.9136,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.656130946570377e-05,
+      "loss": 3.9203,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.655292351819325e-05,
+      "loss": 3.9153,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.654453757068273e-05,
+      "loss": 3.9224,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6536168001975942e-05,
+      "loss": 3.9249,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.652779843326915e-05,
+      "loss": 3.9222,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651941248575863e-05,
+      "loss": 3.9125,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651102653824811e-05,
+      "loss": 3.9106,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6502656969541324e-05,
+      "loss": 3.9246,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6494271022030804e-05,
+      "loss": 3.9068,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6485885074520284e-05,
+      "loss": 3.9247,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6477499127009764e-05,
+      "loss": 3.9128,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6469113179499244e-05,
+      "loss": 3.9187,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6460727231988724e-05,
+      "loss": 3.9162,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6452357663281936e-05,
+      "loss": 3.9157,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6443971715771416e-05,
+      "loss": 3.8992,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6435585768260896e-05,
+      "loss": 3.9137,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6427199820750376e-05,
+      "loss": 3.9045,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.641881387323985e-05,
+      "loss": 3.9188,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.641042792572933e-05,
+      "loss": 3.9336,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6402041978218813e-05,
+      "loss": 3.9264,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6393656030708293e-05,
+      "loss": 3.9002,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6385270083197772e-05,
+      "loss": 3.9283,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.637690051449098e-05,
+      "loss": 3.9178,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636851456698046e-05,
+      "loss": 3.9184,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636012861946994e-05,
+      "loss": 3.9065,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.635174267195942e-05,
+      "loss": 3.9225,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6343356724448905e-05,
+      "loss": 3.9154,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6334970776938385e-05,
+      "loss": 3.9156,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6326584829427865e-05,
+      "loss": 3.9195,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6318198881917345e-05,
+      "loss": 3.9201,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6309829313210554e-05,
+      "loss": 3.9172,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6301443365700034e-05,
+      "loss": 3.9107,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6293057418189514e-05,
+      "loss": 3.9213,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6284671470678997e-05,
+      "loss": 3.919,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6276285523168477e-05,
+      "loss": 3.9126,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6267915954461686e-05,
+      "loss": 3.9255,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6259530006951166e-05,
+      "loss": 3.91,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.020868301391602,
+      "eval_runtime": 295.3362,
+      "eval_samples_per_second": 1292.056,
+      "eval_steps_per_second": 40.378,
+      "step": 1450080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6251144059440646e-05,
+      "loss": 3.9059,
+      "step": 1450496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6242758111930126e-05,
+      "loss": 3.9121,
+      "step": 1451008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6234372164419606e-05,
+      "loss": 3.9253,
+      "step": 1451520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6225986216909086e-05,
+      "loss": 3.9163,
+      "step": 1452032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6217600269398562e-05,
+      "loss": 3.9316,
+      "step": 1452544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6209214321888042e-05,
+      "loss": 3.9082,
+      "step": 1453056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6200828374377522e-05,
+      "loss": 3.9214,
+      "step": 1453568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6192442426867002e-05,
+      "loss": 3.8971,
+      "step": 1454080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6184056479356482e-05,
+      "loss": 3.9276,
+      "step": 1454592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6175670531845962e-05,
+      "loss": 3.9087,
+      "step": 1455104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6167284584335446e-05,
+      "loss": 3.9218,
+      "step": 1455616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6158898636824925e-05,
+      "loss": 3.9207,
+      "step": 1456128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6150529068118135e-05,
+      "loss": 3.9111,
+      "step": 1456640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6142143120607615e-05,
+      "loss": 3.9135,
+      "step": 1457152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6133757173097094e-05,
+      "loss": 3.9082,
+      "step": 1457664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6125371225586574e-05,
+      "loss": 3.9044,
+      "step": 1458176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6116985278076054e-05,
+      "loss": 3.9164,
+      "step": 1458688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6108599330565538e-05,
+      "loss": 3.9136,
+      "step": 1459200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6100213383055018e-05,
+      "loss": 3.9154,
+      "step": 1459712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6091843814348227e-05,
+      "loss": 3.9298,
+      "step": 1460224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6083457866837707e-05,
+      "loss": 3.9283,
+      "step": 1460736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6075071919327187e-05,
+      "loss": 3.9202,
+      "step": 1461248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6066685971816667e-05,
+      "loss": 3.917,
+      "step": 1461760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6058300024306147e-05,
+      "loss": 3.9254,
+      "step": 1462272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.604991407679563e-05,
+      "loss": 3.9095,
+      "step": 1462784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6041528129285103e-05,
+      "loss": 3.926,
+      "step": 1463296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6033142181774583e-05,
+      "loss": 3.9171,
+      "step": 1463808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6024756234264063e-05,
+      "loss": 3.9143,
+      "step": 1464320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6016370286753543e-05,
+      "loss": 3.9101,
+      "step": 1464832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6007984339243023e-05,
+      "loss": 3.9144,
+      "step": 1465344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5999598391732506e-05,
+      "loss": 3.9141,
+      "step": 1465856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5991212444221986e-05,
+      "loss": 3.9206,
+      "step": 1466368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5982826496711466e-05,
+      "loss": 3.9225,
+      "step": 1466880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5974456928004675e-05,
+      "loss": 3.9196,
+      "step": 1467392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5966087359297884e-05,
+      "loss": 3.9176,
+      "step": 1467904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5957701411787368e-05,
+      "loss": 3.9177,
+      "step": 1468416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5949315464276848e-05,
+      "loss": 3.9036,
+      "step": 1468928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5940929516766328e-05,
+      "loss": 3.9157,
+      "step": 1469440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5932543569255808e-05,
+      "loss": 3.9041,
+      "step": 1469952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5924157621745288e-05,
+      "loss": 3.9106,
+      "step": 1470464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5915788053038497e-05,
+      "loss": 3.9125,
+      "step": 1470976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5907402105527977e-05,
+      "loss": 3.9201,
+      "step": 1471488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.589901615801746e-05,
+      "loss": 3.9156,
+      "step": 1472000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.589063021050694e-05,
+      "loss": 3.9069,
+      "step": 1472512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.588224426299642e-05,
+      "loss": 3.9201,
+      "step": 1473024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.58738583154859e-05,
+      "loss": 3.9195,
+      "step": 1473536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.586547236797538e-05,
+      "loss": 3.9207,
+      "step": 1474048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.585710279926859e-05,
+      "loss": 3.9124,
+      "step": 1474560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.584871685175807e-05,
+      "loss": 3.904,
+      "step": 1475072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5840330904247552e-05,
+      "loss": 3.9222,
+      "step": 1475584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5831944956737032e-05,
+      "loss": 3.908,
+      "step": 1476096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5823559009226512e-05,
+      "loss": 3.9189,
+      "step": 1476608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5815173061715992e-05,
+      "loss": 3.9095,
+      "step": 1477120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5806787114205465e-05,
+      "loss": 3.9075,
+      "step": 1477632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5798401166694945e-05,
+      "loss": 3.9063,
+      "step": 1478144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5790047976791894e-05,
+      "loss": 3.912,
+      "step": 1478656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5781662029281374e-05,
+      "loss": 3.9124,
+      "step": 1479168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5773276081770854e-05,
+      "loss": 3.9171,
+      "step": 1479680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5764890134260333e-05,
+      "loss": 3.9142,
+      "step": 1480192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5756504186749813e-05,
+      "loss": 3.9073,
+      "step": 1480704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.574811823923929e-05,
+      "loss": 3.9058,
+      "step": 1481216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.573973229172877e-05,
+      "loss": 3.9264,
+      "step": 1481728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.573134634421825e-05,
+      "loss": 3.8962,
+      "step": 1482240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.572296039670773e-05,
+      "loss": 3.8963,
+      "step": 1482752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.571457444919721e-05,
+      "loss": 3.9093,
+      "step": 1483264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.570620488049042e-05,
+      "loss": 3.9219,
+      "step": 1483776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.56978189329799e-05,
+      "loss": 3.9087,
+      "step": 1484288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.568943298546938e-05,
+      "loss": 3.9043,
+      "step": 1484800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5681047037958862e-05,
+      "loss": 3.8979,
+      "step": 1485312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5672661090448342e-05,
+      "loss": 3.9045,
+      "step": 1485824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.566429152174155e-05,
+      "loss": 3.9214,
+      "step": 1486336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.565590557423103e-05,
+      "loss": 3.9113,
+      "step": 1486848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.564751962672051e-05,
+      "loss": 3.9108,
+      "step": 1487360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.563913367920999e-05,
+      "loss": 3.9205,
+      "step": 1487872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5630764110503204e-05,
+      "loss": 3.9224,
+      "step": 1488384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5622378162992684e-05,
+      "loss": 3.9033,
+      "step": 1488896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5613992215482163e-05,
+      "loss": 3.908,
+      "step": 1489408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5605606267971643e-05,
+      "loss": 3.9195,
+      "step": 1489920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5597236699264853e-05,
+      "loss": 3.8997,
+      "step": 1490432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5588850751754332e-05,
+      "loss": 3.9145,
+      "step": 1490944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5580464804243816e-05,
+      "loss": 3.9028,
+      "step": 1491456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5572078856733296e-05,
+      "loss": 3.9265,
+      "step": 1491968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5563692909222776e-05,
+      "loss": 3.9115,
+      "step": 1492480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5555323340515985e-05,
+      "loss": 3.9029,
+      "step": 1492992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5546937393005465e-05,
+      "loss": 3.9011,
+      "step": 1493504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5538551445494945e-05,
+      "loss": 3.9126,
+      "step": 1494016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5530165497984425e-05,
+      "loss": 3.922,
+      "step": 1494528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5521795929277637e-05,
+      "loss": 3.9115,
+      "step": 1495040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5513426360570846e-05,
+      "loss": 3.9096,
+      "step": 1495552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5505040413060326e-05,
+      "loss": 3.9063,
+      "step": 1496064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5496654465549806e-05,
+      "loss": 3.9163,
+      "step": 1496576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5488268518039286e-05,
+      "loss": 3.9064,
+      "step": 1497088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.547988257052877e-05,
+      "loss": 3.9063,
+      "step": 1497600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.547149662301825e-05,
+      "loss": 3.9183,
+      "step": 1498112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.546312705431146e-05,
+      "loss": 3.9172,
+      "step": 1498624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.545474110680094e-05,
+      "loss": 3.914,
+      "step": 1499136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.544635515929042e-05,
+      "loss": 3.8979,
+      "step": 1499648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.54379692117799e-05,
+      "loss": 3.9122,
+      "step": 1500160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.542958326426938e-05,
+      "loss": 3.9061,
+      "step": 1500672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5421197316758862e-05,
+      "loss": 3.9128,
+      "step": 1501184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.541282774805207e-05,
+      "loss": 3.9115,
+      "step": 1501696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.540444180054155e-05,
+      "loss": 3.904,
+      "step": 1502208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.539607223183476e-05,
+      "loss": 3.9048,
+      "step": 1502720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.538768628432424e-05,
+      "loss": 3.9118,
+      "step": 1503232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5379300336813723e-05,
+      "loss": 3.89,
+      "step": 1503744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5370914389303203e-05,
+      "loss": 3.9092,
+      "step": 1504256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5362544820596412e-05,
+      "loss": 3.8965,
+      "step": 1504768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5354158873085892e-05,
+      "loss": 3.9139,
+      "step": 1505280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5345772925575372e-05,
+      "loss": 3.9063,
+      "step": 1505792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5337386978064852e-05,
+      "loss": 3.9074,
+      "step": 1506304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5329001030554332e-05,
+      "loss": 3.9026,
+      "step": 1506816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5320615083043815e-05,
+      "loss": 3.9094,
+      "step": 1507328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5312229135533295e-05,
+      "loss": 3.9151,
+      "step": 1507840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5303843188022775e-05,
+      "loss": 3.9019,
+      "step": 1508352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.529545724051225e-05,
+      "loss": 3.9131,
+      "step": 1508864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.528707129300173e-05,
+      "loss": 3.9167,
+      "step": 1509376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.527868534549121e-05,
+      "loss": 3.912,
+      "step": 1509888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5270299397980692e-05,
+      "loss": 3.9138,
+      "step": 1510400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.52619298292739e-05,
+      "loss": 3.8978,
+      "step": 1510912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.525354388176338e-05,
+      "loss": 3.9172,
+      "step": 1511424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.524515793425286e-05,
+      "loss": 3.9047,
+      "step": 1511936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.523677198674234e-05,
+      "loss": 3.9123,
+      "step": 1512448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.522838603923182e-05,
+      "loss": 3.905,
+      "step": 1512960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.522003284932877e-05,
+      "loss": 3.9124,
+      "step": 1513472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.521164690181825e-05,
+      "loss": 3.9077,
+      "step": 1513984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5203260954307722e-05,
+      "loss": 3.9036,
+      "step": 1514496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5194875006797202e-05,
+      "loss": 3.8945,
+      "step": 1515008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5186489059286682e-05,
+      "loss": 3.9024,
+      "step": 1515520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5178119490579895e-05,
+      "loss": 3.9017,
+      "step": 1516032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5169733543069375e-05,
+      "loss": 3.9096,
+      "step": 1516544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5161347595558854e-05,
+      "loss": 3.9229,
+      "step": 1517056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5152961648048334e-05,
+      "loss": 3.9191,
+      "step": 1517568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5144575700537814e-05,
+      "loss": 3.8914,
+      "step": 1518080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5136206131831023e-05,
+      "loss": 3.9193,
+      "step": 1518592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5127820184320507e-05,
+      "loss": 3.9088,
+      "step": 1519104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5119434236809987e-05,
+      "loss": 3.9101,
+      "step": 1519616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5111048289299467e-05,
+      "loss": 3.9048,
+      "step": 1520128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5102662341788947e-05,
+      "loss": 3.9108,
+      "step": 1520640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5094276394278427e-05,
+      "loss": 3.9046,
+      "step": 1521152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5085890446767907e-05,
+      "loss": 3.9126,
+      "step": 1521664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5077504499257387e-05,
+      "loss": 3.9078,
+      "step": 1522176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5069118551746867e-05,
+      "loss": 3.9089,
+      "step": 1522688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.506074898304008e-05,
+      "loss": 3.9129,
+      "step": 1523200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.505236303552956e-05,
+      "loss": 3.9025,
+      "step": 1523712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.504397708801904e-05,
+      "loss": 3.9104,
+      "step": 1524224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5035607519312248e-05,
+      "loss": 3.9163,
+      "step": 1524736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5027221571801728e-05,
+      "loss": 3.9042,
+      "step": 1525248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.501885200309494e-05,
+      "loss": 3.915,
+      "step": 1525760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.501046605558442e-05,
+      "loss": 3.9084,
+      "step": 1526272
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.017010688781738,
+      "eval_runtime": 297.0366,
+      "eval_samples_per_second": 1284.66,
+      "eval_steps_per_second": 40.147,
+      "step": 1526400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.50020801080739e-05,
+      "loss": 3.9043,
+      "step": 1526784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.499369416056338e-05,
+      "loss": 3.9038,
+      "step": 1527296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.498530821305286e-05,
+      "loss": 3.9131,
+      "step": 1527808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.497692226554234e-05,
+      "loss": 3.9104,
+      "step": 1528320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4968536318031817e-05,
+      "loss": 3.9246,
+      "step": 1528832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.49601503705213e-05,
+      "loss": 3.8987,
+      "step": 1529344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4951780801814513e-05,
+      "loss": 3.9113,
+      "step": 1529856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.494339485430399e-05,
+      "loss": 3.8948,
+      "step": 1530368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.493500890679347e-05,
+      "loss": 3.9142,
+      "step": 1530880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.492662295928295e-05,
+      "loss": 3.9048,
+      "step": 1531392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.491823701177243e-05,
+      "loss": 3.9097,
+      "step": 1531904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.490986744306564e-05,
+      "loss": 3.916,
+      "step": 1532416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4901497874358854e-05,
+      "loss": 3.9008,
+      "step": 1532928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4893111926848334e-05,
+      "loss": 3.9081,
+      "step": 1533440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4884725979337814e-05,
+      "loss": 3.9005,
+      "step": 1533952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.487634003182729e-05,
+      "loss": 3.8966,
+      "step": 1534464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.486795408431677e-05,
+      "loss": 3.9056,
+      "step": 1534976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4859568136806254e-05,
+      "loss": 3.9059,
+      "step": 1535488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4851198568099463e-05,
+      "loss": 3.9071,
+      "step": 1536000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4842812620588943e-05,
+      "loss": 3.9193,
+      "step": 1536512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4834426673078423e-05,
+      "loss": 3.9209,
+      "step": 1537024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4826040725567903e-05,
+      "loss": 3.9135,
+      "step": 1537536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4817654778057383e-05,
+      "loss": 3.9083,
+      "step": 1538048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4809268830546863e-05,
+      "loss": 3.9173,
+      "step": 1538560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4800882883036343e-05,
+      "loss": 3.8982,
+      "step": 1539072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4792496935525823e-05,
+      "loss": 3.9157,
+      "step": 1539584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4784110988015303e-05,
+      "loss": 3.9101,
+      "step": 1540096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4775725040504783e-05,
+      "loss": 3.9037,
+      "step": 1540608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4767339092994262e-05,
+      "loss": 3.9039,
+      "step": 1541120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.475896952428747e-05,
+      "loss": 3.9037,
+      "step": 1541632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4750583576776955e-05,
+      "loss": 3.9086,
+      "step": 1542144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4742197629266435e-05,
+      "loss": 3.9104,
+      "step": 1542656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4733811681755915e-05,
+      "loss": 3.9146,
+      "step": 1543168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4725425734245395e-05,
+      "loss": 3.9099,
+      "step": 1543680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4717072544342336e-05,
+      "loss": 3.9116,
+      "step": 1544192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4708686596831816e-05,
+      "loss": 3.9096,
+      "step": 1544704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4700300649321296e-05,
+      "loss": 3.8943,
+      "step": 1545216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4691914701810776e-05,
+      "loss": 3.9092,
+      "step": 1545728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4683528754300256e-05,
+      "loss": 3.8964,
+      "step": 1546240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4675142806789736e-05,
+      "loss": 3.9004,
+      "step": 1546752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4666756859279216e-05,
+      "loss": 3.902,
+      "step": 1547264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4658370911768696e-05,
+      "loss": 3.9173,
+      "step": 1547776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4649984964258176e-05,
+      "loss": 3.9046,
+      "step": 1548288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4641599016747656e-05,
+      "loss": 3.898,
+      "step": 1548800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.463322944804087e-05,
+      "loss": 3.9134,
+      "step": 1549312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.462484350053035e-05,
+      "loss": 3.907,
+      "step": 1549824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4616457553019825e-05,
+      "loss": 3.9117,
+      "step": 1550336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4608087984313037e-05,
+      "loss": 3.9071,
+      "step": 1550848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4599702036802517e-05,
+      "loss": 3.8972,
+      "step": 1551360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4591316089292e-05,
+      "loss": 3.9147,
+      "step": 1551872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4582930141781477e-05,
+      "loss": 3.9005,
+      "step": 1552384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4574544194270957e-05,
+      "loss": 3.9115,
+      "step": 1552896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4566158246760437e-05,
+      "loss": 3.9021,
+      "step": 1553408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4557772299249917e-05,
+      "loss": 3.8986,
+      "step": 1553920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4549386351739397e-05,
+      "loss": 3.9006,
+      "step": 1554432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.454101678303261e-05,
+      "loss": 3.9048,
+      "step": 1554944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.453263083552209e-05,
+      "loss": 3.9039,
+      "step": 1555456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.452424488801157e-05,
+      "loss": 3.9063,
+      "step": 1555968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.451585894050105e-05,
+      "loss": 3.9122,
+      "step": 1556480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.450748937179426e-05,
+      "loss": 3.9011,
+      "step": 1556992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.449910342428374e-05,
+      "loss": 3.895,
+      "step": 1557504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.449071747677322e-05,
+      "loss": 3.9221,
+      "step": 1558016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4482331529262702e-05,
+      "loss": 3.8867,
+      "step": 1558528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4473945581752182e-05,
+      "loss": 3.8871,
+      "step": 1559040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.446557601304539e-05,
+      "loss": 3.9021,
+      "step": 1559552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.445719006553487e-05,
+      "loss": 3.9175,
+      "step": 1560064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.444880411802435e-05,
+      "loss": 3.897,
+      "step": 1560576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4440418170513827e-05,
+      "loss": 3.9003,
+      "step": 1561088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.443203222300331e-05,
+      "loss": 3.8917,
+      "step": 1561600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.442364627549279e-05,
+      "loss": 3.8983,
+      "step": 1562112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.441526032798227e-05,
+      "loss": 3.9097,
+      "step": 1562624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.440687438047175e-05,
+      "loss": 3.9047,
+      "step": 1563136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.439848843296123e-05,
+      "loss": 3.9044,
+      "step": 1563648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4390135243058172e-05,
+      "loss": 3.9084,
+      "step": 1564160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4381749295547656e-05,
+      "loss": 3.9141,
+      "step": 1564672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4373363348037132e-05,
+      "loss": 3.8996,
+      "step": 1565184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4364977400526612e-05,
+      "loss": 3.9003,
+      "step": 1565696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4356591453016092e-05,
+      "loss": 3.9135,
+      "step": 1566208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4348205505505572e-05,
+      "loss": 3.8904,
+      "step": 1566720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4339819557995052e-05,
+      "loss": 3.9062,
+      "step": 1567232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4331433610484532e-05,
+      "loss": 3.8965,
+      "step": 1567744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4323064041777744e-05,
+      "loss": 3.9186,
+      "step": 1568256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4314678094267224e-05,
+      "loss": 3.9043,
+      "step": 1568768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4306292146756704e-05,
+      "loss": 3.8961,
+      "step": 1569280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4297906199246184e-05,
+      "loss": 3.8881,
+      "step": 1569792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.428952025173566e-05,
+      "loss": 3.9081,
+      "step": 1570304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.428113430422514e-05,
+      "loss": 3.9104,
+      "step": 1570816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4272764735518357e-05,
+      "loss": 3.9077,
+      "step": 1571328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4264395166811566e-05,
+      "loss": 3.9002,
+      "step": 1571840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4256009219301046e-05,
+      "loss": 3.902,
+      "step": 1572352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4247623271790526e-05,
+      "loss": 3.9074,
+      "step": 1572864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4239237324280006e-05,
+      "loss": 3.8978,
+      "step": 1573376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4230851376769486e-05,
+      "loss": 3.8979,
+      "step": 1573888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4222465429258966e-05,
+      "loss": 3.9121,
+      "step": 1574400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4214095860552178e-05,
+      "loss": 3.9085,
+      "step": 1574912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4205709913041658e-05,
+      "loss": 3.9065,
+      "step": 1575424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4197323965531135e-05,
+      "loss": 3.8941,
+      "step": 1575936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4188938018020614e-05,
+      "loss": 3.9052,
+      "step": 1576448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4180552070510094e-05,
+      "loss": 3.8961,
+      "step": 1576960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4172166122999574e-05,
+      "loss": 3.9073,
+      "step": 1577472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4163780175489058e-05,
+      "loss": 3.9025,
+      "step": 1577984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4155394227978538e-05,
+      "loss": 3.9006,
+      "step": 1578496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4147008280468018e-05,
+      "loss": 3.8963,
+      "step": 1579008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4138638711761227e-05,
+      "loss": 3.8977,
+      "step": 1579520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4130252764250707e-05,
+      "loss": 3.8898,
+      "step": 1580032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4121866816740187e-05,
+      "loss": 3.8961,
+      "step": 1580544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4113480869229667e-05,
+      "loss": 3.8873,
+      "step": 1581056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.410511130052288e-05,
+      "loss": 3.9057,
+      "step": 1581568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.409672535301236e-05,
+      "loss": 3.9008,
+      "step": 1582080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.408833940550184e-05,
+      "loss": 3.8976,
+      "step": 1582592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4079953457991316e-05,
+      "loss": 3.8948,
+      "step": 1583104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4071567510480796e-05,
+      "loss": 3.9063,
+      "step": 1583616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.406318156297028e-05,
+      "loss": 3.9044,
+      "step": 1584128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.405481199426349e-05,
+      "loss": 3.8938,
+      "step": 1584640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4046426046752968e-05,
+      "loss": 3.9063,
+      "step": 1585152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4038040099242448e-05,
+      "loss": 3.9094,
+      "step": 1585664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4029654151731928e-05,
+      "loss": 3.9047,
+      "step": 1586176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4021268204221408e-05,
+      "loss": 3.9046,
+      "step": 1586688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4012882256710888e-05,
+      "loss": 3.8918,
+      "step": 1587200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.400449630920037e-05,
+      "loss": 3.9068,
+      "step": 1587712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3996110361689848e-05,
+      "loss": 3.8986,
+      "step": 1588224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3987724414179328e-05,
+      "loss": 3.9054,
+      "step": 1588736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.397935484547254e-05,
+      "loss": 3.8897,
+      "step": 1589248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.397096889796202e-05,
+      "loss": 3.9052,
+      "step": 1589760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3962582950451497e-05,
+      "loss": 3.9009,
+      "step": 1590272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.395419700294098e-05,
+      "loss": 3.8964,
+      "step": 1590784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.394581105543046e-05,
+      "loss": 3.8889,
+      "step": 1591296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.393744148672367e-05,
+      "loss": 3.8936,
+      "step": 1591808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.392905553921315e-05,
+      "loss": 3.8918,
+      "step": 1592320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.392068597050636e-05,
+      "loss": 3.9057,
+      "step": 1592832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.391230002299584e-05,
+      "loss": 3.9098,
+      "step": 1593344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.390391407548532e-05,
+      "loss": 3.9138,
+      "step": 1593856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.38955281279748e-05,
+      "loss": 3.8852,
+      "step": 1594368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.388714218046428e-05,
+      "loss": 3.9119,
+      "step": 1594880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.387875623295376e-05,
+      "loss": 3.9041,
+      "step": 1595392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.387037028544324e-05,
+      "loss": 3.8994,
+      "step": 1595904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.386198433793272e-05,
+      "loss": 3.8943,
+      "step": 1596416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3853614769225934e-05,
+      "loss": 3.9081,
+      "step": 1596928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3845228821715414e-05,
+      "loss": 3.8993,
+      "step": 1597440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3836842874204894e-05,
+      "loss": 3.9044,
+      "step": 1597952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3828456926694374e-05,
+      "loss": 3.8965,
+      "step": 1598464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.382007097918385e-05,
+      "loss": 3.9,
+      "step": 1598976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3811701410477063e-05,
+      "loss": 3.9041,
+      "step": 1599488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3803315462966543e-05,
+      "loss": 3.8976,
+      "step": 1600000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3794945894259755e-05,
+      "loss": 3.9017,
+      "step": 1600512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3786559946749235e-05,
+      "loss": 3.9087,
+      "step": 1601024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3778173999238715e-05,
+      "loss": 3.8986,
+      "step": 1601536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3769804430531924e-05,
+      "loss": 3.9033,
+      "step": 1602048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3761418483021404e-05,
+      "loss": 3.9049,
+      "step": 1602560
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0136871337890625,
+      "eval_runtime": 292.691,
+      "eval_samples_per_second": 1303.733,
+      "eval_steps_per_second": 40.743,
+      "step": 1602720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3753032535510887e-05,
+      "loss": 3.8952,
+      "step": 1603072
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3744646588000367e-05,
+      "loss": 3.895,
+      "step": 1603584
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3736260640489847e-05,
+      "loss": 3.906,
+      "step": 1604096
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3727874692979324e-05,
+      "loss": 3.9047,
+      "step": 1604608
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3719488745468804e-05,
+      "loss": 3.913,
+      "step": 1605120
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3711102797958284e-05,
+      "loss": 3.8959,
+      "step": 1605632
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3702716850447764e-05,
+      "loss": 3.9029,
+      "step": 1606144
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3694347281740976e-05,
+      "loss": 3.8897,
+      "step": 1606656
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3685961334230456e-05,
+      "loss": 3.904,
+      "step": 1607168
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3677575386719936e-05,
+      "loss": 3.8979,
+      "step": 1607680
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3669189439209416e-05,
+      "loss": 3.9011,
+      "step": 1608192
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3660803491698896e-05,
+      "loss": 3.9064,
+      "step": 1608704
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3652433922992105e-05,
+      "loss": 3.8982,
+      "step": 1609216
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.364406435428532e-05,
+      "loss": 3.8973,
+      "step": 1609728
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3635678406774797e-05,
+      "loss": 3.8962,
+      "step": 1610240
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3627292459264277e-05,
+      "loss": 3.8885,
+      "step": 1610752
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3618906511753757e-05,
+      "loss": 3.8939,
+      "step": 1611264
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3610520564243237e-05,
+      "loss": 3.9021,
+      "step": 1611776
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3602134616732717e-05,
+      "loss": 3.9001,
+      "step": 1612288
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3593748669222197e-05,
+      "loss": 3.9151,
+      "step": 1612800
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.358536272171168e-05,
+      "loss": 3.9122,
+      "step": 1613312
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.357699315300489e-05,
+      "loss": 3.9078,
+      "step": 1613824
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.356860720549437e-05,
+      "loss": 3.9016,
+      "step": 1614336
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.356022125798385e-05,
+      "loss": 3.9095,
+      "step": 1614848
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.355183531047333e-05,
+      "loss": 3.8966,
+      "step": 1615360
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.3543449362962806e-05,
+      "loss": 3.9083,
+      "step": 1615872
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.353506341545229e-05,
+      "loss": 3.9029,
+      "step": 1616384
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.352667746794177e-05,
+      "loss": 3.8952,
+      "step": 1616896
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.351829152043125e-05,
+      "loss": 3.895,
+      "step": 1617408
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.350992195172446e-05,
+      "loss": 3.8998,
+      "step": 1617920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.350155238301767e-05,
+      "loss": 3.9003,
+      "step": 1618432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.349316643550715e-05,
+      "loss": 3.9039,
+      "step": 1618944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.348478048799663e-05,
+      "loss": 3.9094,
+      "step": 1619456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.347639454048611e-05,
+      "loss": 3.9029,
+      "step": 1619968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3468024971779323e-05,
+      "loss": 3.905,
+      "step": 1620480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3459639024268803e-05,
+      "loss": 3.904,
+      "step": 1620992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.345125307675828e-05,
+      "loss": 3.8919,
+      "step": 1621504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.344286712924776e-05,
+      "loss": 3.8951,
+      "step": 1622016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3434481181737243e-05,
+      "loss": 3.8906,
+      "step": 1622528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3426095234226723e-05,
+      "loss": 3.8925,
+      "step": 1623040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3417709286716203e-05,
+      "loss": 3.8953,
+      "step": 1623552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3409323339205683e-05,
+      "loss": 3.9111,
+      "step": 1624064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.340093739169516e-05,
+      "loss": 3.9005,
+      "step": 1624576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.339255144418464e-05,
+      "loss": 3.8912,
+      "step": 1625088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3384198254281585e-05,
+      "loss": 3.9094,
+      "step": 1625600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3375812306771065e-05,
+      "loss": 3.8952,
+      "step": 1626112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3367426359260544e-05,
+      "loss": 3.9098,
+      "step": 1626624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3359040411750024e-05,
+      "loss": 3.8965,
+      "step": 1627136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3350654464239504e-05,
+      "loss": 3.8875,
+      "step": 1627648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.334226851672898e-05,
+      "loss": 3.9065,
+      "step": 1628160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3333898948022197e-05,
+      "loss": 3.8942,
+      "step": 1628672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3325513000511677e-05,
+      "loss": 3.9074,
+      "step": 1629184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3317127053001157e-05,
+      "loss": 3.8901,
+      "step": 1629696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3308741105490633e-05,
+      "loss": 3.8947,
+      "step": 1630208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3300355157980113e-05,
+      "loss": 3.8897,
+      "step": 1630720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3291985589273326e-05,
+      "loss": 3.8981,
+      "step": 1631232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3283599641762806e-05,
+      "loss": 3.8993,
+      "step": 1631744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3275213694252286e-05,
+      "loss": 3.8952,
+      "step": 1632256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3266827746741766e-05,
+      "loss": 3.9057,
+      "step": 1632768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3258441799231246e-05,
+      "loss": 3.8957,
+      "step": 1633280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3250055851720726e-05,
+      "loss": 3.8879,
+      "step": 1633792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3241669904210205e-05,
+      "loss": 3.9165,
+      "step": 1634304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3233283956699685e-05,
+      "loss": 3.879,
+      "step": 1634816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3224898009189165e-05,
+      "loss": 3.8781,
+      "step": 1635328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3216528440482378e-05,
+      "loss": 3.8946,
+      "step": 1635840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3208142492971858e-05,
+      "loss": 3.9116,
+      "step": 1636352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3199756545461338e-05,
+      "loss": 3.887,
+      "step": 1636864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3191370597950814e-05,
+      "loss": 3.8966,
+      "step": 1637376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3183001029244027e-05,
+      "loss": 3.8822,
+      "step": 1637888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3174615081733507e-05,
+      "loss": 3.8896,
+      "step": 1638400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.316622913422299e-05,
+      "loss": 3.9031,
+      "step": 1638912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3157843186712467e-05,
+      "loss": 3.8997,
+      "step": 1639424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.314947361800568e-05,
+      "loss": 3.8991,
+      "step": 1639936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.314108767049516e-05,
+      "loss": 3.9001,
+      "step": 1640448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3132718101788368e-05,
+      "loss": 3.9074,
+      "step": 1640960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.312433215427785e-05,
+      "loss": 3.8873,
+      "step": 1641472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.311594620676733e-05,
+      "loss": 3.8967,
+      "step": 1641984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.310756025925681e-05,
+      "loss": 3.9022,
+      "step": 1642496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3099174311746288e-05,
+      "loss": 3.886,
+      "step": 1643008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3090788364235768e-05,
+      "loss": 3.896,
+      "step": 1643520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3082402416725248e-05,
+      "loss": 3.8937,
+      "step": 1644032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.307403284801846e-05,
+      "loss": 3.907,
+      "step": 1644544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.306564690050794e-05,
+      "loss": 3.9007,
+      "step": 1645056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.305726095299742e-05,
+      "loss": 3.8891,
+      "step": 1645568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.30488750054869e-05,
+      "loss": 3.8824,
+      "step": 1646080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.304048905797638e-05,
+      "loss": 3.8976,
+      "step": 1646592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.303210311046586e-05,
+      "loss": 3.9017,
+      "step": 1647104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.302371716295534e-05,
+      "loss": 3.8968,
+      "step": 1647616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.3015347594248553e-05,
+      "loss": 3.8978,
+      "step": 1648128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3006961646738033e-05,
+      "loss": 3.8956,
+      "step": 1648640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2998575699227513e-05,
+      "loss": 3.8998,
+      "step": 1649152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2990189751716993e-05,
+      "loss": 3.8902,
+      "step": 1649664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.298180380420647e-05,
+      "loss": 3.8894,
+      "step": 1650176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.297341785669595e-05,
+      "loss": 3.9074,
+      "step": 1650688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.296503190918543e-05,
+      "loss": 3.9016,
+      "step": 1651200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2956645961674912e-05,
+      "loss": 3.8975,
+      "step": 1651712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.294827639296812e-05,
+      "loss": 3.8872,
+      "step": 1652224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2939906824261334e-05,
+      "loss": 3.8996,
+      "step": 1652736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2931520876750814e-05,
+      "loss": 3.8857,
+      "step": 1653248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.292313492924029e-05,
+      "loss": 3.8994,
+      "step": 1653760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2914765360533506e-05,
+      "loss": 3.8964,
+      "step": 1654272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2906379413022986e-05,
+      "loss": 3.8945,
+      "step": 1654784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2897993465512466e-05,
+      "loss": 3.8901,
+      "step": 1655296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2889607518001943e-05,
+      "loss": 3.8861,
+      "step": 1655808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2881221570491423e-05,
+      "loss": 3.892,
+      "step": 1656320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2872835622980903e-05,
+      "loss": 3.8859,
+      "step": 1656832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2864449675470383e-05,
+      "loss": 3.8858,
+      "step": 1657344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2856063727959866e-05,
+      "loss": 3.8963,
+      "step": 1657856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2847677780449346e-05,
+      "loss": 3.8964,
+      "step": 1658368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2839291832938826e-05,
+      "loss": 3.8907,
+      "step": 1658880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2830922264232035e-05,
+      "loss": 3.8847,
+      "step": 1659392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2822536316721515e-05,
+      "loss": 3.8958,
+      "step": 1659904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2814150369210995e-05,
+      "loss": 3.9006,
+      "step": 1660416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2805764421700475e-05,
+      "loss": 3.8919,
+      "step": 1660928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2797378474189955e-05,
+      "loss": 3.9002,
+      "step": 1661440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2789008905483167e-05,
+      "loss": 3.8999,
+      "step": 1661952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2780622957972647e-05,
+      "loss": 3.8992,
+      "step": 1662464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2772237010462124e-05,
+      "loss": 3.8951,
+      "step": 1662976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2763851062951604e-05,
+      "loss": 3.8874,
+      "step": 1663488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2755465115441084e-05,
+      "loss": 3.895,
+      "step": 1664000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2747079167930567e-05,
+      "loss": 3.895,
+      "step": 1664512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2738693220420047e-05,
+      "loss": 3.9004,
+      "step": 1665024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2730323651713256e-05,
+      "loss": 3.8869,
+      "step": 1665536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2721937704202736e-05,
+      "loss": 3.8974,
+      "step": 1666048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2713551756692216e-05,
+      "loss": 3.8955,
+      "step": 1666560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2705165809181696e-05,
+      "loss": 3.8914,
+      "step": 1667072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2696779861671176e-05,
+      "loss": 3.8775,
+      "step": 1667584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2688393914160656e-05,
+      "loss": 3.891,
+      "step": 1668096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.268002434545387e-05,
+      "loss": 3.8843,
+      "step": 1668608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.267163839794335e-05,
+      "loss": 3.8993,
+      "step": 1669120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2663268829236557e-05,
+      "loss": 3.9033,
+      "step": 1669632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2654882881726037e-05,
+      "loss": 3.9068,
+      "step": 1670144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.264649693421552e-05,
+      "loss": 3.8801,
+      "step": 1670656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2638110986705e-05,
+      "loss": 3.906,
+      "step": 1671168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2629725039194477e-05,
+      "loss": 3.8966,
+      "step": 1671680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2621339091683957e-05,
+      "loss": 3.8928,
+      "step": 1672192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2612953144173437e-05,
+      "loss": 3.8878,
+      "step": 1672704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2604567196662917e-05,
+      "loss": 3.9025,
+      "step": 1673216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2596181249152397e-05,
+      "loss": 3.8884,
+      "step": 1673728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2587795301641877e-05,
+      "loss": 3.9003,
+      "step": 1674240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.257940935413136e-05,
+      "loss": 3.8879,
+      "step": 1674752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.2571023406620837e-05,
+      "loss": 3.8974,
+      "step": 1675264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.256265383791405e-05,
+      "loss": 3.8958,
+      "step": 1675776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.255426789040353e-05,
+      "loss": 3.8919,
+      "step": 1676288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.254588194289301e-05,
+      "loss": 3.8997,
+      "step": 1676800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.253749599538249e-05,
+      "loss": 3.8948,
+      "step": 1677312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.252911004787197e-05,
+      "loss": 3.8985,
+      "step": 1677824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.252075685796891e-05,
+      "loss": 3.8899,
+      "step": 1678336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.251237091045839e-05,
+      "loss": 3.9022,
+      "step": 1678848
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.011056423187256,
+      "eval_runtime": 294.1966,
+      "eval_samples_per_second": 1297.061,
+      "eval_steps_per_second": 40.534,
+      "step": 1679040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.250398496294787e-05,
+      "loss": 3.893,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.249559901543735e-05,
+      "loss": 3.8859,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.248721306792683e-05,
+      "loss": 3.897,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247882712041631e-05,
+      "loss": 3.9001,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247044117290579e-05,
+      "loss": 3.9053,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.246205522539527e-05,
+      "loss": 3.8893,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.245366927788475e-05,
+      "loss": 3.893,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.244528333037423e-05,
+      "loss": 3.8845,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.243689738286371e-05,
+      "loss": 3.8992,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.242851143535319e-05,
+      "loss": 3.8909,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.242012548784267e-05,
+      "loss": 3.8974,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.241173954033215e-05,
+      "loss": 3.8977,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2403369971625363e-05,
+      "loss": 3.8905,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.239498402411484e-05,
+      "loss": 3.8933,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.238659807660432e-05,
+      "loss": 3.8841,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.23782121290938e-05,
+      "loss": 3.8814,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2369826181583283e-05,
+      "loss": 3.8871,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2361440234072763e-05,
+      "loss": 3.8988,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2353054286562243e-05,
+      "loss": 3.8895,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2344668339051723e-05,
+      "loss": 3.9107,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.23362823915412e-05,
+      "loss": 3.9025,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.232789644403068e-05,
+      "loss": 3.903,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231951049652016e-05,
+      "loss": 3.8942,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231112454900964e-05,
+      "loss": 3.9002,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.230275498030285e-05,
+      "loss": 3.8865,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.229436903279233e-05,
+      "loss": 3.9051,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2285999464085544e-05,
+      "loss": 3.8933,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.227761351657502e-05,
+      "loss": 3.8884,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.22692275690645e-05,
+      "loss": 3.892,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2260841621553984e-05,
+      "loss": 3.8931,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2252455674043464e-05,
+      "loss": 3.8934,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2244069726532944e-05,
+      "loss": 3.9003,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2235683779022424e-05,
+      "loss": 3.897,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2227297831511904e-05,
+      "loss": 3.9017,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2218928262805113e-05,
+      "loss": 3.8909,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2210558694098325e-05,
+      "loss": 3.897,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2202189125391538e-05,
+      "loss": 3.8914,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2193803177881018e-05,
+      "loss": 3.8841,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2185417230370494e-05,
+      "loss": 3.8829,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2177031282859974e-05,
+      "loss": 3.8877,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2168645335349454e-05,
+      "loss": 3.8887,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2160259387838937e-05,
+      "loss": 3.904,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2151873440328417e-05,
+      "loss": 3.8929,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2143487492817897e-05,
+      "loss": 3.8862,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2135101545307377e-05,
+      "loss": 3.905,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2126731976600586e-05,
+      "loss": 3.8883,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2118346029090066e-05,
+      "loss": 3.9027,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2109960081579546e-05,
+      "loss": 3.892,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2101574134069026e-05,
+      "loss": 3.8753,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2093188186558506e-05,
+      "loss": 3.9064,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2084802239047986e-05,
+      "loss": 3.8853,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2076416291537466e-05,
+      "loss": 3.9009,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2068030344026946e-05,
+      "loss": 3.8887,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2059660775320155e-05,
+      "loss": 3.888,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205127482780964e-05,
+      "loss": 3.8807,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.204290525910285e-05,
+      "loss": 3.8891,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2034519311592328e-05,
+      "loss": 3.8996,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2026133364081808e-05,
+      "loss": 3.8844,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2017747416571287e-05,
+      "loss": 3.8989,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2009361469060767e-05,
+      "loss": 3.8927,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2000975521550247e-05,
+      "loss": 3.882,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.199258957403973e-05,
+      "loss": 3.9071,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1984203626529207e-05,
+      "loss": 3.8749,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.197583405782242e-05,
+      "loss": 3.8718,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.19674481103119e-05,
+      "loss": 3.8911,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.195906216280138e-05,
+      "loss": 3.9009,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1950676215290856e-05,
+      "loss": 3.8801,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1942306646584072e-05,
+      "loss": 3.8935,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1933920699073552e-05,
+      "loss": 3.8764,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1925534751563032e-05,
+      "loss": 3.89,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.191716518285624e-05,
+      "loss": 3.8939,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.190877923534572e-05,
+      "loss": 3.8889,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.19003932878352e-05,
+      "loss": 3.8939,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.189200734032468e-05,
+      "loss": 3.8906,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.188362139281416e-05,
+      "loss": 3.9023,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.187523544530364e-05,
+      "loss": 3.8821,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.186684949779312e-05,
+      "loss": 3.8937,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.18584635502826e-05,
+      "loss": 3.8894,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.185009398157581e-05,
+      "loss": 3.8838,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1841724412869026e-05,
+      "loss": 3.8884,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1833338465358506e-05,
+      "loss": 3.889,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1824952517847982e-05,
+      "loss": 3.8994,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1816566570337462e-05,
+      "loss": 3.8995,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1808197001630675e-05,
+      "loss": 3.8777,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1799811054120155e-05,
+      "loss": 3.8781,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1791425106609635e-05,
+      "loss": 3.8919,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1783039159099115e-05,
+      "loss": 3.8974,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1774653211588595e-05,
+      "loss": 3.8921,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1766283642881804e-05,
+      "loss": 3.8851,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1757897695371284e-05,
+      "loss": 3.8932,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1749511747860764e-05,
+      "loss": 3.89,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1741125800350247e-05,
+      "loss": 3.8875,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1732739852839727e-05,
+      "loss": 3.8812,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1724370284132936e-05,
+      "loss": 3.8979,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.171600071542615e-05,
+      "loss": 3.8938,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1707614767915625e-05,
+      "loss": 3.8955,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169922882040511e-05,
+      "loss": 3.8813,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169084287289459e-05,
+      "loss": 3.8904,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1682456925384068e-05,
+      "loss": 3.8816,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1674070977873548e-05,
+      "loss": 3.8934,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1665701409166757e-05,
+      "loss": 3.888,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1657315461656237e-05,
+      "loss": 3.8838,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1648929514145717e-05,
+      "loss": 3.8887,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.16405435666352e-05,
+      "loss": 3.8786,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.163215761912468e-05,
+      "loss": 3.8842,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.162377167161416e-05,
+      "loss": 3.8789,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1615385724103637e-05,
+      "loss": 3.8828,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1606999776593117e-05,
+      "loss": 3.8827,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1598613829082597e-05,
+      "loss": 3.8963,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1590227881572077e-05,
+      "loss": 3.8781,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1581841934061557e-05,
+      "loss": 3.885,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.157345598655104e-05,
+      "loss": 3.8849,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.156508641784425e-05,
+      "loss": 3.892,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.155670047033373e-05,
+      "loss": 3.8841,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.154831452282321e-05,
+      "loss": 3.8981,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.153992857531269e-05,
+      "loss": 3.8937,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1531559006605902e-05,
+      "loss": 3.893,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.152318943789911e-05,
+      "loss": 3.8897,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.151480349038859e-05,
+      "loss": 3.886,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.150641754287807e-05,
+      "loss": 3.8866,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.149803159536755e-05,
+      "loss": 3.8927,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1489662026660763e-05,
+      "loss": 3.8883,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1481276079150243e-05,
+      "loss": 3.8826,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1472906510443456e-05,
+      "loss": 3.8928,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1464520562932932e-05,
+      "loss": 3.889,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1456134615422412e-05,
+      "loss": 3.8863,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1447748667911892e-05,
+      "loss": 3.8705,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1439362720401372e-05,
+      "loss": 3.8825,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1430976772890855e-05,
+      "loss": 3.8772,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1422590825380335e-05,
+      "loss": 3.892,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1414204877869812e-05,
+      "loss": 3.8969,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1405835309163024e-05,
+      "loss": 3.8989,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1397449361652504e-05,
+      "loss": 3.8749,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1389063414141984e-05,
+      "loss": 3.8979,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1380677466631464e-05,
+      "loss": 3.8903,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1372291519120944e-05,
+      "loss": 3.8891,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1363905571610424e-05,
+      "loss": 3.8839,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1355519624099904e-05,
+      "loss": 3.8959,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1347150055393113e-05,
+      "loss": 3.883,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1338764107882593e-05,
+      "loss": 3.8986,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1330378160372077e-05,
+      "loss": 3.8789,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1322008591665286e-05,
+      "loss": 3.89,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1313622644154766e-05,
+      "loss": 3.8891,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1305236696644246e-05,
+      "loss": 3.8873,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1296850749133725e-05,
+      "loss": 3.8905,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1288464801623205e-05,
+      "loss": 3.8887,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1280078854112685e-05,
+      "loss": 3.8915,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1271709285405898e-05,
+      "loss": 3.8846,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1263323337895378e-05,
+      "loss": 3.8974,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.008915901184082,
+      "eval_runtime": 294.4439,
+      "eval_samples_per_second": 1295.972,
+      "eval_steps_per_second": 40.5,
+      "step": 1755360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1254937390384858e-05,
+      "loss": 3.8839,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1246551442874338e-05,
+      "loss": 3.8778,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1238165495363818e-05,
+      "loss": 3.8924,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1229779547853294e-05,
+      "loss": 3.894,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1221393600342778e-05,
+      "loss": 3.8968,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1213007652832258e-05,
+      "loss": 3.8848,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1204621705321738e-05,
+      "loss": 3.8886,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1196235757811217e-05,
+      "loss": 3.8771,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1187849810300697e-05,
+      "loss": 3.8897,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1179463862790174e-05,
+      "loss": 3.8886,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1171094294083386e-05,
+      "loss": 3.889,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.116270834657287e-05,
+      "loss": 3.8921,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.115433877786608e-05,
+      "loss": 3.8894,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.114595283035556e-05,
+      "loss": 3.8826,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.113756688284504e-05,
+      "loss": 3.8792,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.112918093533452e-05,
+      "loss": 3.8737,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1120794987824e-05,
+      "loss": 3.8838,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.111240904031348e-05,
+      "loss": 3.8912,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.110402309280296e-05,
+      "loss": 3.8831,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.109563714529244e-05,
+      "loss": 3.8998,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.108725119778192e-05,
+      "loss": 3.8988,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1078881629075128e-05,
+      "loss": 3.8989,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.107051206036834e-05,
+      "loss": 3.8871,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.106212611285782e-05,
+      "loss": 3.8898,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.10537401653473e-05,
+      "loss": 3.887,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.104535421783678e-05,
+      "loss": 3.8952,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.103696827032626e-05,
+      "loss": 3.8862,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.102858232281574e-05,
+      "loss": 3.8867,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.102019637530522e-05,
+      "loss": 3.8877,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.10118104277947e-05,
+      "loss": 3.8846,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.100342448028418e-05,
+      "loss": 3.8832,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.099503853277366e-05,
+      "loss": 3.8931,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.098665258526314e-05,
+      "loss": 3.8926,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.097826663775262e-05,
+      "loss": 3.8939,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.096991344784956e-05,
+      "loss": 3.8817,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.096152750033904e-05,
+      "loss": 3.8964,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0953141552828525e-05,
+      "loss": 3.8877,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0944755605318005e-05,
+      "loss": 3.8747,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.093636965780748e-05,
+      "loss": 3.8768,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.092798371029696e-05,
+      "loss": 3.8816,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0919614141590174e-05,
+      "loss": 3.8853,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.091122819407965e-05,
+      "loss": 3.8926,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0902858625372866e-05,
+      "loss": 3.8881,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0894472677862346e-05,
+      "loss": 3.8778,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0886086730351826e-05,
+      "loss": 3.9019,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0877700782841302e-05,
+      "loss": 3.8818,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0869314835330782e-05,
+      "loss": 3.8973,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0860928887820262e-05,
+      "loss": 3.8889,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0852542940309742e-05,
+      "loss": 3.8636,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0844156992799226e-05,
+      "loss": 3.9049,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0835771045288706e-05,
+      "loss": 3.8776,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0827385097778186e-05,
+      "loss": 3.8917,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0819015529071395e-05,
+      "loss": 3.8837,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0810629581560875e-05,
+      "loss": 3.8798,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0802243634050355e-05,
+      "loss": 3.8778,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0793857686539835e-05,
+      "loss": 3.8788,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0785471739029314e-05,
+      "loss": 3.8992,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0777102170322527e-05,
+      "loss": 3.8728,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0768716222812007e-05,
+      "loss": 3.8928,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0760330275301484e-05,
+      "loss": 3.8889,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0751944327790963e-05,
+      "loss": 3.8711,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0743558380280443e-05,
+      "loss": 3.8951,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0735172432769927e-05,
+      "loss": 3.8751,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0726786485259407e-05,
+      "loss": 3.8693,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0718416916552616e-05,
+      "loss": 3.8813,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0710030969042096e-05,
+      "loss": 3.8933,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0701645021531576e-05,
+      "loss": 3.8749,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0693259074021056e-05,
+      "loss": 3.8876,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0684873126510536e-05,
+      "loss": 3.8708,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0676487179000016e-05,
+      "loss": 3.8791,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0668101231489496e-05,
+      "loss": 3.8903,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0659715283978975e-05,
+      "loss": 3.8816,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0651329336468455e-05,
+      "loss": 3.8915,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0642959767761665e-05,
+      "loss": 3.8832,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0634573820251148e-05,
+      "loss": 3.8951,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0626187872740628e-05,
+      "loss": 3.8792,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0617801925230108e-05,
+      "loss": 3.8867,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0609432356523317e-05,
+      "loss": 3.8805,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0601046409012797e-05,
+      "loss": 3.8783,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.059267684030601e-05,
+      "loss": 3.8832,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.058429089279549e-05,
+      "loss": 3.8829,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.057590494528497e-05,
+      "loss": 3.8921,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.056751899777445e-05,
+      "loss": 3.8969,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055913305026393e-05,
+      "loss": 3.8676,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055074710275341e-05,
+      "loss": 3.872,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.054236115524289e-05,
+      "loss": 3.8848,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.053397520773237e-05,
+      "loss": 3.8907,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.052560563902558e-05,
+      "loss": 3.8864,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.051723607031879e-05,
+      "loss": 3.8797,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.050885012280827e-05,
+      "loss": 3.8921,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.050046417529775e-05,
+      "loss": 3.8827,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.049207822778723e-05,
+      "loss": 3.8832,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.048369228027671e-05,
+      "loss": 3.8766,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.047530633276619e-05,
+      "loss": 3.8876,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.046692038525567e-05,
+      "loss": 3.8915,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.045853443774515e-05,
+      "loss": 3.8894,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0450164869038363e-05,
+      "loss": 3.8783,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0441778921527843e-05,
+      "loss": 3.8812,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.043339297401732e-05,
+      "loss": 3.8762,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0425023405310535e-05,
+      "loss": 3.8888,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0416637457800015e-05,
+      "loss": 3.8831,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0408251510289492e-05,
+      "loss": 3.8782,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.039986556277897e-05,
+      "loss": 3.8828,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.039147961526845e-05,
+      "loss": 3.8746,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.038309366775793e-05,
+      "loss": 3.8753,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.037470772024741e-05,
+      "loss": 3.8704,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0366338151540624e-05,
+      "loss": 3.8794,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0357952204030104e-05,
+      "loss": 3.8779,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0349566256519584e-05,
+      "loss": 3.8874,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0341180309009064e-05,
+      "loss": 3.8779,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0332794361498544e-05,
+      "loss": 3.8698,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0324408413988024e-05,
+      "loss": 3.8822,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0316022466477504e-05,
+      "loss": 3.8839,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0307636518966984e-05,
+      "loss": 3.8813,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0299266950260196e-05,
+      "loss": 3.8888,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0290881002749673e-05,
+      "loss": 3.8837,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0282495055239153e-05,
+      "loss": 3.8934,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0274109107728633e-05,
+      "loss": 3.8796,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0265723160218113e-05,
+      "loss": 3.8778,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0257337212707596e-05,
+      "loss": 3.8785,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0248951265197076e-05,
+      "loss": 3.8905,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0240565317686556e-05,
+      "loss": 3.8793,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0232212127783498e-05,
+      "loss": 3.8749,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.022384255907671e-05,
+      "loss": 3.8852,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.021545661156619e-05,
+      "loss": 3.8887,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.020707066405567e-05,
+      "loss": 3.8765,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0198684716545146e-05,
+      "loss": 3.8691,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0190298769034626e-05,
+      "loss": 3.8756,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0181912821524106e-05,
+      "loss": 3.873,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0173526874013586e-05,
+      "loss": 3.886,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0165140926503066e-05,
+      "loss": 3.8912,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.015677135779628e-05,
+      "loss": 3.8932,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.014838541028576e-05,
+      "loss": 3.8664,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013999946277524e-05,
+      "loss": 3.8941,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013161351526472e-05,
+      "loss": 3.8836,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.01232275677542e-05,
+      "loss": 3.8826,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.011484162024368e-05,
+      "loss": 3.8771,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.010647205153689e-05,
+      "loss": 3.8865,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.009808610402637e-05,
+      "loss": 3.8775,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.008970015651585e-05,
+      "loss": 3.8938,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0081314209005328e-05,
+      "loss": 3.875,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.007294464029854e-05,
+      "loss": 3.8813,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.006455869278802e-05,
+      "loss": 3.884,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0056172745277503e-05,
+      "loss": 3.8844,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.004778679776698e-05,
+      "loss": 3.8811,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.003940085025646e-05,
+      "loss": 3.8816,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.003101490274594e-05,
+      "loss": 3.8873,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0022645334039152e-05,
+      "loss": 3.881,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.001425938652863e-05,
+      "loss": 3.8887,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.006490230560303,
+      "eval_runtime": 292.2164,
+      "eval_samples_per_second": 1305.851,
+      "eval_steps_per_second": 40.809,
+      "step": 1831680
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.0005873439018112e-05,
+      "loss": 3.8743,
+      "step": 1831936
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9997487491507592e-05,
+      "loss": 3.8738,
+      "step": 1832448
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9989101543997072e-05,
+      "loss": 3.8843,
+      "step": 1832960
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.998073197529028e-05,
+      "loss": 3.8859,
+      "step": 1833472
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.997234602777976e-05,
+      "loss": 3.8932,
+      "step": 1833984
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.996396008026924e-05,
+      "loss": 3.8786,
+      "step": 1834496
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.995557413275872e-05,
+      "loss": 3.8868,
+      "step": 1835008
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9947188185248204e-05,
+      "loss": 3.8755,
+      "step": 1835520
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9938802237737684e-05,
+      "loss": 3.8781,
+      "step": 1836032
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.993041629022716e-05,
+      "loss": 3.8816,
+      "step": 1836544
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.992203034271664e-05,
+      "loss": 3.8814,
+      "step": 1837056
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9913660774009853e-05,
+      "loss": 3.8861,
+      "step": 1837568
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9905291205303066e-05,
+      "loss": 3.8881,
+      "step": 1838080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9896905257792546e-05,
+      "loss": 3.8741,
+      "step": 1838592
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9888519310282026e-05,
+      "loss": 3.8765,
+      "step": 1839104
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9880133362771506e-05,
+      "loss": 3.8678,
+      "step": 1839616
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9871747415260982e-05,
+      "loss": 3.8781,
+      "step": 1840128
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9863361467750462e-05,
+      "loss": 3.8892,
+      "step": 1840640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9854991899043675e-05,
+      "loss": 3.8735,
+      "step": 1841152
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9846622330336887e-05,
+      "loss": 3.8921,
+      "step": 1841664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9838236382826367e-05,
+      "loss": 3.8923,
+      "step": 1842176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9829850435315847e-05,
+      "loss": 3.8903,
+      "step": 1842688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9821464487805327e-05,
+      "loss": 3.889,
+      "step": 1843200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9813094919098536e-05,
+      "loss": 3.8797,
+      "step": 1843712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.980470897158802e-05,
+      "loss": 3.8831,
+      "step": 1844224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.97963230240775e-05,
+      "loss": 3.8871,
+      "step": 1844736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.978793707656698e-05,
+      "loss": 3.8876,
+      "step": 1845248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9779551129056456e-05,
+      "loss": 3.88,
+      "step": 1845760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9771165181545936e-05,
+      "loss": 3.8811,
+      "step": 1846272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9762779234035416e-05,
+      "loss": 3.8809,
+      "step": 1846784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9754393286524896e-05,
+      "loss": 3.8758,
+      "step": 1847296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9746007339014376e-05,
+      "loss": 3.8863,
+      "step": 1847808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.973762139150386e-05,
+      "loss": 3.8842,
+      "step": 1848320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9729251822797068e-05,
+      "loss": 3.8899,
+      "step": 1848832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9720865875286548e-05,
+      "loss": 3.8775,
+      "step": 1849344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9712496306579757e-05,
+      "loss": 3.8895,
+      "step": 1849856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9704110359069237e-05,
+      "loss": 3.8829,
+      "step": 1850368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.969572441155872e-05,
+      "loss": 3.8677,
+      "step": 1850880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.96873384640482e-05,
+      "loss": 3.8773,
+      "step": 1851392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.967895251653768e-05,
+      "loss": 3.8715,
+      "step": 1851904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.967056656902716e-05,
+      "loss": 3.8764,
+      "step": 1852416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9662180621516637e-05,
+      "loss": 3.8922,
+      "step": 1852928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9653794674006117e-05,
+      "loss": 3.882,
+      "step": 1853440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9645408726495597e-05,
+      "loss": 3.8761,
+      "step": 1853952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9637039157788813e-05,
+      "loss": 3.8876,
+      "step": 1854464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.962865321027829e-05,
+      "loss": 3.8763,
+      "step": 1854976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.962026726276777e-05,
+      "loss": 3.8941,
+      "step": 1855488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.961188131525725e-05,
+      "loss": 3.8813,
+      "step": 1856000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.960349536774673e-05,
+      "loss": 3.8588,
+      "step": 1856512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.959510942023621e-05,
+      "loss": 3.896,
+      "step": 1857024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.958672347272569e-05,
+      "loss": 3.8734,
+      "step": 1857536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.957833752521517e-05,
+      "loss": 3.8861,
+      "step": 1858048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.956996795650838e-05,
+      "loss": 3.8781,
+      "step": 1858560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.956158200899786e-05,
+      "loss": 3.8735,
+      "step": 1859072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.955319606148734e-05,
+      "loss": 3.8722,
+      "step": 1859584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9544810113976818e-05,
+      "loss": 3.8749,
+      "step": 1860096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.953644054527003e-05,
+      "loss": 3.8954,
+      "step": 1860608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9528054597759514e-05,
+      "loss": 3.865,
+      "step": 1861120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9519668650248994e-05,
+      "loss": 3.8901,
+      "step": 1861632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9511299081542203e-05,
+      "loss": 3.8809,
+      "step": 1862144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9502913134031683e-05,
+      "loss": 3.8635,
+      "step": 1862656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9494527186521163e-05,
+      "loss": 3.8928,
+      "step": 1863168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9486141239010643e-05,
+      "loss": 3.8724,
+      "step": 1863680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9477755291500123e-05,
+      "loss": 3.8597,
+      "step": 1864192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9469369343989603e-05,
+      "loss": 3.8755,
+      "step": 1864704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9460999775282815e-05,
+      "loss": 3.8859,
+      "step": 1865216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9452613827772292e-05,
+      "loss": 3.8722,
+      "step": 1865728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9444227880261772e-05,
+      "loss": 3.8799,
+      "step": 1866240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9435841932751252e-05,
+      "loss": 3.8687,
+      "step": 1866752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9427472364044468e-05,
+      "loss": 3.8678,
+      "step": 1867264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9419086416533944e-05,
+      "loss": 3.8833,
+      "step": 1867776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9410700469023424e-05,
+      "loss": 3.8791,
+      "step": 1868288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9402314521512904e-05,
+      "loss": 3.8818,
+      "step": 1868800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9393928574002384e-05,
+      "loss": 3.8786,
+      "step": 1869312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9385542626491864e-05,
+      "loss": 3.8915,
+      "step": 1869824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9377156678981344e-05,
+      "loss": 3.8731,
+      "step": 1870336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9368787110274556e-05,
+      "loss": 3.881,
+      "step": 1870848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9360401162764036e-05,
+      "loss": 3.8731,
+      "step": 1871360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9352015215253516e-05,
+      "loss": 3.8745,
+      "step": 1871872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9343629267742996e-05,
+      "loss": 3.8781,
+      "step": 1872384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9335243320232473e-05,
+      "loss": 3.8732,
+      "step": 1872896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9326857372721953e-05,
+      "loss": 3.8861,
+      "step": 1873408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9318471425211436e-05,
+      "loss": 3.8925,
+      "step": 1873920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9310085477700916e-05,
+      "loss": 3.8621,
+      "step": 1874432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9301715908994125e-05,
+      "loss": 3.8672,
+      "step": 1874944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9293329961483605e-05,
+      "loss": 3.8793,
+      "step": 1875456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9284944013973085e-05,
+      "loss": 3.8806,
+      "step": 1875968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9276558066462565e-05,
+      "loss": 3.8823,
+      "step": 1876480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9268188497755778e-05,
+      "loss": 3.8766,
+      "step": 1876992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.925981892904899e-05,
+      "loss": 3.8818,
+      "step": 1877504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.925143298153847e-05,
+      "loss": 3.8783,
+      "step": 1878016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9243047034027947e-05,
+      "loss": 3.8744,
+      "step": 1878528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9234661086517427e-05,
+      "loss": 3.8724,
+      "step": 1879040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9226275139006906e-05,
+      "loss": 3.8833,
+      "step": 1879552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.921788919149639e-05,
+      "loss": 3.8871,
+      "step": 1880064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.920950324398587e-05,
+      "loss": 3.8821,
+      "step": 1880576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.920111729647535e-05,
+      "loss": 3.8703,
+      "step": 1881088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9192731348964826e-05,
+      "loss": 3.878,
+      "step": 1881600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.918436178025804e-05,
+      "loss": 3.8702,
+      "step": 1882112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.917597583274752e-05,
+      "loss": 3.8798,
+      "step": 1882624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.916760626404073e-05,
+      "loss": 3.8789,
+      "step": 1883136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.915922031653021e-05,
+      "loss": 3.8683,
+      "step": 1883648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.915083436901969e-05,
+      "loss": 3.8803,
+      "step": 1884160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.914244842150917e-05,
+      "loss": 3.8653,
+      "step": 1884672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.913406247399865e-05,
+      "loss": 3.8721,
+      "step": 1885184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9125676526488128e-05,
+      "loss": 3.865,
+      "step": 1885696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9117306957781343e-05,
+      "loss": 3.8759,
+      "step": 1886208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9108921010270823e-05,
+      "loss": 3.8726,
+      "step": 1886720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.91005350627603e-05,
+      "loss": 3.8802,
+      "step": 1887232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.909214911524978e-05,
+      "loss": 3.8681,
+      "step": 1887744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.908376316773926e-05,
+      "loss": 3.8696,
+      "step": 1888256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.907537722022874e-05,
+      "loss": 3.8721,
+      "step": 1888768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.906699127271822e-05,
+      "loss": 3.8812,
+      "step": 1889280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.90586053252077e-05,
+      "loss": 3.8735,
+      "step": 1889792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9050235756500912e-05,
+      "loss": 3.8798,
+      "step": 1890304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9041849808990392e-05,
+      "loss": 3.8773,
+      "step": 1890816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9033463861479872e-05,
+      "loss": 3.8898,
+      "step": 1891328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.902509429277308e-05,
+      "loss": 3.8765,
+      "step": 1891840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.901670834526256e-05,
+      "loss": 3.8725,
+      "step": 1892352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9008322397752045e-05,
+      "loss": 3.8761,
+      "step": 1892864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8999936450241525e-05,
+      "loss": 3.8818,
+      "step": 1893376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8991566881534734e-05,
+      "loss": 3.8761,
+      "step": 1893888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8983180934024214e-05,
+      "loss": 3.8704,
+      "step": 1894400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8974794986513694e-05,
+      "loss": 3.8772,
+      "step": 1894912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8966409039003173e-05,
+      "loss": 3.882,
+      "step": 1895424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8958023091492653e-05,
+      "loss": 3.8707,
+      "step": 1895936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8949637143982133e-05,
+      "loss": 3.8613,
+      "step": 1896448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8941267575275346e-05,
+      "loss": 3.8701,
+      "step": 1896960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8932881627764826e-05,
+      "loss": 3.8668,
+      "step": 1897472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8924495680254306e-05,
+      "loss": 3.8832,
+      "step": 1897984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8916109732743782e-05,
+      "loss": 3.8808,
+      "step": 1898496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8907723785233262e-05,
+      "loss": 3.8872,
+      "step": 1899008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8899337837722746e-05,
+      "loss": 3.8615,
+      "step": 1899520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8890968269015955e-05,
+      "loss": 3.8882,
+      "step": 1900032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8882582321505435e-05,
+      "loss": 3.8789,
+      "step": 1900544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8874196373994915e-05,
+      "loss": 3.8776,
+      "step": 1901056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8865810426484395e-05,
+      "loss": 3.8719,
+      "step": 1901568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8857424478973875e-05,
+      "loss": 3.8758,
+      "step": 1902080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8849038531463355e-05,
+      "loss": 3.8735,
+      "step": 1902592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8840652583952838e-05,
+      "loss": 3.888,
+      "step": 1903104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.883229939404978e-05,
+      "loss": 3.8666,
+      "step": 1903616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8823913446539256e-05,
+      "loss": 3.8778,
+      "step": 1904128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8815527499028736e-05,
+      "loss": 3.8815,
+      "step": 1904640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8807141551518216e-05,
+      "loss": 3.8817,
+      "step": 1905152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.87987556040077e-05,
+      "loss": 3.87,
+      "step": 1905664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.879036965649718e-05,
+      "loss": 3.8764,
+      "step": 1906176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.878198370898666e-05,
+      "loss": 3.8798,
+      "step": 1906688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8773597761476136e-05,
+      "loss": 3.8759,
+      "step": 1907200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8765211813965616e-05,
+      "loss": 3.8836,
+      "step": 1907712
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.004377841949463,
+      "eval_runtime": 305.833,
+      "eval_samples_per_second": 1247.71,
+      "eval_steps_per_second": 38.992,
+      "step": 1908000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8756825866455096e-05,
+      "loss": 3.8752,
+      "step": 1908224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8748439918944576e-05,
+      "loss": 3.8669,
+      "step": 1908736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8740053971434056e-05,
+      "loss": 3.8761,
+      "step": 1909248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.873166802392354e-05,
+      "loss": 3.8825,
+      "step": 1909760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.872328207641302e-05,
+      "loss": 3.8889,
+      "step": 1910272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8714896128902495e-05,
+      "loss": 3.8736,
+      "step": 1910784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8706510181391975e-05,
+      "loss": 3.88,
+      "step": 1911296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8698124233881455e-05,
+      "loss": 3.8663,
+      "step": 1911808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8689738286370935e-05,
+      "loss": 3.8748,
+      "step": 1912320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8681352338860415e-05,
+      "loss": 3.8769,
+      "step": 1912832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.86729663913499e-05,
+      "loss": 3.876,
+      "step": 1913344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.866458044383938e-05,
+      "loss": 3.8802,
+      "step": 1913856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8656210875132588e-05,
+      "loss": 3.8862,
+      "step": 1914368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8647824927622068e-05,
+      "loss": 3.8672,
+      "step": 1914880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8639438980111548e-05,
+      "loss": 3.868,
+      "step": 1915392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8631053032601024e-05,
+      "loss": 3.8583,
+      "step": 1915904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8622667085090508e-05,
+      "loss": 3.8757,
+      "step": 1916416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8614281137579987e-05,
+      "loss": 3.8839,
+      "step": 1916928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8605895190069467e-05,
+      "loss": 3.8664,
+      "step": 1917440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8597509242558947e-05,
+      "loss": 3.8837,
+      "step": 1917952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8589123295048427e-05,
+      "loss": 3.8898,
+      "step": 1918464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8580753726341636e-05,
+      "loss": 3.8848,
+      "step": 1918976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.857238415763485e-05,
+      "loss": 3.8818,
+      "step": 1919488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.856399821012433e-05,
+      "loss": 3.8745,
+      "step": 1920000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.855561226261381e-05,
+      "loss": 3.8801,
+      "step": 1920512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.854722631510329e-05,
+      "loss": 3.878,
+      "step": 1921024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.853884036759277e-05,
+      "loss": 3.8816,
+      "step": 1921536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.853045442008225e-05,
+      "loss": 3.881,
+      "step": 1922048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.852206847257173e-05,
+      "loss": 3.8733,
+      "step": 1922560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.851368252506121e-05,
+      "loss": 3.8681,
+      "step": 1923072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.850529657755069e-05,
+      "loss": 3.8708,
+      "step": 1923584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.849691063004017e-05,
+      "loss": 3.8842,
+      "step": 1924096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.848852468252965e-05,
+      "loss": 3.8829,
+      "step": 1924608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.848013873501913e-05,
+      "loss": 3.8821,
+      "step": 1925120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.847175278750861e-05,
+      "loss": 3.8725,
+      "step": 1925632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8463399597605553e-05,
+      "loss": 3.8781,
+      "step": 1926144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.845501365009503e-05,
+      "loss": 3.8791,
+      "step": 1926656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.844662770258451e-05,
+      "loss": 3.8635,
+      "step": 1927168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.843824175507399e-05,
+      "loss": 3.8697,
+      "step": 1927680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.842985580756347e-05,
+      "loss": 3.8688,
+      "step": 1928192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.842146986005295e-05,
+      "loss": 3.8699,
+      "step": 1928704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.841308391254243e-05,
+      "loss": 3.8856,
+      "step": 1929216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.840469796503191e-05,
+      "loss": 3.8759,
+      "step": 1929728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8396328396325122e-05,
+      "loss": 3.8704,
+      "step": 1930240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8387942448814602e-05,
+      "loss": 3.8843,
+      "step": 1930752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8379556501304082e-05,
+      "loss": 3.868,
+      "step": 1931264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8371170553793562e-05,
+      "loss": 3.8813,
+      "step": 1931776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.836280098508677e-05,
+      "loss": 3.885,
+      "step": 1932288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8354415037576255e-05,
+      "loss": 3.8521,
+      "step": 1932800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8346029090065734e-05,
+      "loss": 3.8868,
+      "step": 1933312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.833764314255521e-05,
+      "loss": 3.8704,
+      "step": 1933824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.832925719504469e-05,
+      "loss": 3.8807,
+      "step": 1934336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.832087124753417e-05,
+      "loss": 3.8732,
+      "step": 1934848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.831248530002365e-05,
+      "loss": 3.8677,
+      "step": 1935360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8304115731316863e-05,
+      "loss": 3.8687,
+      "step": 1935872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8295729783806343e-05,
+      "loss": 3.8676,
+      "step": 1936384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8287343836295823e-05,
+      "loss": 3.8909,
+      "step": 1936896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8278957888785303e-05,
+      "loss": 3.859,
+      "step": 1937408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8270571941274783e-05,
+      "loss": 3.8858,
+      "step": 1937920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8262185993764263e-05,
+      "loss": 3.8761,
+      "step": 1938432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8253800046253743e-05,
+      "loss": 3.8592,
+      "step": 1938944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8245430477546956e-05,
+      "loss": 3.8859,
+      "step": 1939456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8237044530036436e-05,
+      "loss": 3.8659,
+      "step": 1939968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8228658582525916e-05,
+      "loss": 3.8561,
+      "step": 1940480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8220272635015392e-05,
+      "loss": 3.8654,
+      "step": 1940992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8211886687504872e-05,
+      "loss": 3.8855,
+      "step": 1941504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8203500739994352e-05,
+      "loss": 3.8634,
+      "step": 1942016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8195114792483832e-05,
+      "loss": 3.8779,
+      "step": 1942528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8186745223777044e-05,
+      "loss": 3.8648,
+      "step": 1943040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8178359276266524e-05,
+      "loss": 3.8615,
+      "step": 1943552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8169973328756004e-05,
+      "loss": 3.8775,
+      "step": 1944064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8161587381245484e-05,
+      "loss": 3.8738,
+      "step": 1944576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8153201433734964e-05,
+      "loss": 3.8778,
+      "step": 1945088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8144815486224444e-05,
+      "loss": 3.8748,
+      "step": 1945600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8136429538713924e-05,
+      "loss": 3.881,
+      "step": 1946112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8128043591203404e-05,
+      "loss": 3.8722,
+      "step": 1946624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8119674022496617e-05,
+      "loss": 3.8761,
+      "step": 1947136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8111288074986097e-05,
+      "loss": 3.8628,
+      "step": 1947648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8102902127475573e-05,
+      "loss": 3.87,
+      "step": 1948160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8094516179965053e-05,
+      "loss": 3.873,
+      "step": 1948672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8086146611258266e-05,
+      "loss": 3.8662,
+      "step": 1949184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8077777042551478e-05,
+      "loss": 3.8804,
+      "step": 1949696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8069391095040958e-05,
+      "loss": 3.8838,
+      "step": 1950208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8061005147530438e-05,
+      "loss": 3.8654,
+      "step": 1950720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8052619200019918e-05,
+      "loss": 3.8594,
+      "step": 1951232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8044233252509394e-05,
+      "loss": 3.8735,
+      "step": 1951744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8035847304998878e-05,
+      "loss": 3.8753,
+      "step": 1952256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8027461357488358e-05,
+      "loss": 3.876,
+      "step": 1952768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8019075409977838e-05,
+      "loss": 3.8684,
+      "step": 1953280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8010689462467318e-05,
+      "loss": 3.8807,
+      "step": 1953792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8002303514956798e-05,
+      "loss": 3.8715,
+      "step": 1954304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7993933946250007e-05,
+      "loss": 3.8711,
+      "step": 1954816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7985547998739487e-05,
+      "loss": 3.8683,
+      "step": 1955328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.797716205122897e-05,
+      "loss": 3.8753,
+      "step": 1955840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.796877610371845e-05,
+      "loss": 3.8784,
+      "step": 1956352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.796039015620793e-05,
+      "loss": 3.8785,
+      "step": 1956864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7952004208697407e-05,
+      "loss": 3.8623,
+      "step": 1957376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.794363463999062e-05,
+      "loss": 3.8746,
+      "step": 1957888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.79352486924801e-05,
+      "loss": 3.8635,
+      "step": 1958400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.792687912377331e-05,
+      "loss": 3.8775,
+      "step": 1958912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.791849317626279e-05,
+      "loss": 3.873,
+      "step": 1959424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.791010722875227e-05,
+      "loss": 3.8665,
+      "step": 1959936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.790172128124175e-05,
+      "loss": 3.8747,
+      "step": 1960448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7893335333731228e-05,
+      "loss": 3.8578,
+      "step": 1960960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7884949386220708e-05,
+      "loss": 3.8666,
+      "step": 1961472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7876563438710188e-05,
+      "loss": 3.8611,
+      "step": 1961984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.786817749119967e-05,
+      "loss": 3.87,
+      "step": 1962496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.785979154368915e-05,
+      "loss": 3.8678,
+      "step": 1963008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.785142197498236e-05,
+      "loss": 3.8732,
+      "step": 1963520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.784303602747184e-05,
+      "loss": 3.8612,
+      "step": 1964032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.783465007996132e-05,
+      "loss": 3.8637,
+      "step": 1964544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.78262641324508e-05,
+      "loss": 3.8669,
+      "step": 1965056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.781787818494028e-05,
+      "loss": 3.8751,
+      "step": 1965568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.780949223742976e-05,
+      "loss": 3.8695,
+      "step": 1966080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7801122668722972e-05,
+      "loss": 3.8728,
+      "step": 1966592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7792736721212452e-05,
+      "loss": 3.8723,
+      "step": 1967104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7784350773701932e-05,
+      "loss": 3.8813,
+      "step": 1967616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.777598120499514e-05,
+      "loss": 3.8704,
+      "step": 1968128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7767595257484625e-05,
+      "loss": 3.8689,
+      "step": 1968640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7759209309974105e-05,
+      "loss": 3.87,
+      "step": 1969152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7750823362463585e-05,
+      "loss": 3.8755,
+      "step": 1969664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.774243741495306e-05,
+      "loss": 3.872,
+      "step": 1970176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.773405146744254e-05,
+      "loss": 3.8655,
+      "step": 1970688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.772566551993202e-05,
+      "loss": 3.8679,
+      "step": 1971200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.77172795724215e-05,
+      "loss": 3.8773,
+      "step": 1971712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.770889362491098e-05,
+      "loss": 3.8657,
+      "step": 1972224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7700507677400464e-05,
+      "loss": 3.8574,
+      "step": 1972736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.769212172988994e-05,
+      "loss": 3.8625,
+      "step": 1973248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.768373578237942e-05,
+      "loss": 3.8587,
+      "step": 1973760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7675366213672633e-05,
+      "loss": 3.8826,
+      "step": 1974272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7666980266162113e-05,
+      "loss": 3.8702,
+      "step": 1974784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7658610697455326e-05,
+      "loss": 3.8849,
+      "step": 1975296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7650224749944806e-05,
+      "loss": 3.8609,
+      "step": 1975808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7641838802434286e-05,
+      "loss": 3.8772,
+      "step": 1976320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7633452854923766e-05,
+      "loss": 3.873,
+      "step": 1976832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7625066907413242e-05,
+      "loss": 3.8738,
+      "step": 1977344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7616680959902722e-05,
+      "loss": 3.8707,
+      "step": 1977856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7608311391195935e-05,
+      "loss": 3.8676,
+      "step": 1978368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7599925443685415e-05,
+      "loss": 3.866,
+      "step": 1978880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7591539496174895e-05,
+      "loss": 3.8858,
+      "step": 1979392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7583153548664375e-05,
+      "loss": 3.8617,
+      "step": 1979904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7574767601153855e-05,
+      "loss": 3.8723,
+      "step": 1980416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7566381653643335e-05,
+      "loss": 3.8772,
+      "step": 1980928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7557995706132815e-05,
+      "loss": 3.8696,
+      "step": 1981440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7549609758622294e-05,
+      "loss": 3.8657,
+      "step": 1981952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7541240189915507e-05,
+      "loss": 3.8709,
+      "step": 1982464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7532854242404987e-05,
+      "loss": 3.8748,
+      "step": 1982976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7524484673698196e-05,
+      "loss": 3.8739,
+      "step": 1983488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7516098726187676e-05,
+      "loss": 3.8786,
+      "step": 1984000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.002153396606445,
+      "eval_runtime": 315.6877,
+      "eval_samples_per_second": 1208.761,
+      "eval_steps_per_second": 37.775,
+      "step": 1984320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7507712778677156e-05,
+      "loss": 3.8613,
+      "step": 1984512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7499326831166636e-05,
+      "loss": 3.8633,
+      "step": 1985024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.749094088365612e-05,
+      "loss": 3.8745,
+      "step": 1985536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7482554936145596e-05,
+      "loss": 3.8764,
+      "step": 1986048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7474168988635076e-05,
+      "loss": 3.8837,
+      "step": 1986560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7465783041124556e-05,
+      "loss": 3.8694,
+      "step": 1987072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7457397093614036e-05,
+      "loss": 3.8735,
+      "step": 1987584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7449011146103516e-05,
+      "loss": 3.865,
+      "step": 1988096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7440625198592996e-05,
+      "loss": 3.8681,
+      "step": 1988608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7432255629886208e-05,
+      "loss": 3.8721,
+      "step": 1989120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7423869682375688e-05,
+      "loss": 3.8692,
+      "step": 1989632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7415483734865168e-05,
+      "loss": 3.8784,
+      "step": 1990144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7407114166158377e-05,
+      "loss": 3.88,
+      "step": 1990656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7398728218647857e-05,
+      "loss": 3.8624,
+      "step": 1991168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.739034227113734e-05,
+      "loss": 3.8654,
+      "step": 1991680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.738195632362682e-05,
+      "loss": 3.8548,
+      "step": 1992192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.73735703761163e-05,
+      "loss": 3.874,
+      "step": 1992704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7365184428605777e-05,
+      "loss": 3.8741,
+      "step": 1993216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7356798481095257e-05,
+      "loss": 3.8624,
+      "step": 1993728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7348412533584737e-05,
+      "loss": 3.8746,
+      "step": 1994240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7340026586074217e-05,
+      "loss": 3.8896,
+      "step": 1994752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.733165701736743e-05,
+      "loss": 3.8779,
+      "step": 1995264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.732327106985691e-05,
+      "loss": 3.8775,
+      "step": 1995776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.731488512234639e-05,
+      "loss": 3.8696,
+      "step": 1996288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.730649917483587e-05,
+      "loss": 3.8779,
+      "step": 1996800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.729811322732535e-05,
+      "loss": 3.872,
+      "step": 1997312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.728972727981483e-05,
+      "loss": 3.872,
+      "step": 1997824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.728135771110804e-05,
+      "loss": 3.8755,
+      "step": 1998336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.727297176359752e-05,
+      "loss": 3.8708,
+      "step": 1998848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7264585816087e-05,
+      "loss": 3.8665,
+      "step": 1999360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.725619986857648e-05,
+      "loss": 3.8669,
+      "step": 1999872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7247813921065958e-05,
+      "loss": 3.876,
+      "step": 2000384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7239427973555438e-05,
+      "loss": 3.8737,
+      "step": 2000896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.723105840484865e-05,
+      "loss": 3.8796,
+      "step": 2001408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7222672457338134e-05,
+      "loss": 3.8692,
+      "step": 2001920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7214302888631343e-05,
+      "loss": 3.8725,
+      "step": 2002432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7205916941120823e-05,
+      "loss": 3.8755,
+      "step": 2002944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7197530993610303e-05,
+      "loss": 3.8613,
+      "step": 2003456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.718914504609978e-05,
+      "loss": 3.8624,
+      "step": 2003968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.718075909858926e-05,
+      "loss": 3.864,
+      "step": 2004480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7172373151078743e-05,
+      "loss": 3.862,
+      "step": 2004992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7163987203568222e-05,
+      "loss": 3.8769,
+      "step": 2005504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7155601256057702e-05,
+      "loss": 3.8705,
+      "step": 2006016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7147215308547182e-05,
+      "loss": 3.8711,
+      "step": 2006528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7138829361036662e-05,
+      "loss": 3.8712,
+      "step": 2007040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.713044341352614e-05,
+      "loss": 3.8671,
+      "step": 2007552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.712205746601562e-05,
+      "loss": 3.8784,
+      "step": 2008064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7113687897308835e-05,
+      "loss": 3.8779,
+      "step": 2008576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7105301949798315e-05,
+      "loss": 3.8459,
+      "step": 2009088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.709691600228779e-05,
+      "loss": 3.8816,
+      "step": 2009600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.708853005477727e-05,
+      "loss": 3.8632,
+      "step": 2010112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7080160486070484e-05,
+      "loss": 3.878,
+      "step": 2010624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7071774538559964e-05,
+      "loss": 3.872,
+      "step": 2011136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7063388591049444e-05,
+      "loss": 3.8588,
+      "step": 2011648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7055019022342656e-05,
+      "loss": 3.8685,
+      "step": 2012160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7046633074832136e-05,
+      "loss": 3.859,
+      "step": 2012672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7038247127321613e-05,
+      "loss": 3.882,
+      "step": 2013184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7029861179811093e-05,
+      "loss": 3.8581,
+      "step": 2013696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7021475232300573e-05,
+      "loss": 3.876,
+      "step": 2014208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7013089284790053e-05,
+      "loss": 3.8714,
+      "step": 2014720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7004703337279536e-05,
+      "loss": 3.8604,
+      "step": 2015232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6996317389769016e-05,
+      "loss": 3.8793,
+      "step": 2015744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6987964199865957e-05,
+      "loss": 3.8596,
+      "step": 2016256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6979578252355434e-05,
+      "loss": 3.8536,
+      "step": 2016768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6971192304844914e-05,
+      "loss": 3.8572,
+      "step": 2017280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6962806357334397e-05,
+      "loss": 3.8779,
+      "step": 2017792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6954420409823877e-05,
+      "loss": 3.8588,
+      "step": 2018304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6946034462313357e-05,
+      "loss": 3.8766,
+      "step": 2018816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6937648514802837e-05,
+      "loss": 3.8539,
+      "step": 2019328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6929262567292317e-05,
+      "loss": 3.8608,
+      "step": 2019840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.692090937738926e-05,
+      "loss": 3.8695,
+      "step": 2020352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.691252342987874e-05,
+      "loss": 3.8699,
+      "step": 2020864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.690413748236822e-05,
+      "loss": 3.8727,
+      "step": 2021376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.68957515348577e-05,
+      "loss": 3.8698,
+      "step": 2021888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.688736558734718e-05,
+      "loss": 3.8767,
+      "step": 2022400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.687897963983666e-05,
+      "loss": 3.8634,
+      "step": 2022912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.687059369232614e-05,
+      "loss": 3.8694,
+      "step": 2023424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.686222412361935e-05,
+      "loss": 3.8606,
+      "step": 2023936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.685383817610883e-05,
+      "loss": 3.866,
+      "step": 2024448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.684545222859831e-05,
+      "loss": 3.868,
+      "step": 2024960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.683706628108779e-05,
+      "loss": 3.8594,
+      "step": 2025472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6828680333577267e-05,
+      "loss": 3.875,
+      "step": 2025984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6820294386066747e-05,
+      "loss": 3.8784,
+      "step": 2026496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6811908438556227e-05,
+      "loss": 3.8651,
+      "step": 2027008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.680353886984944e-05,
+      "loss": 3.8494,
+      "step": 2027520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.679515292233892e-05,
+      "loss": 3.8707,
+      "step": 2028032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.67867669748284e-05,
+      "loss": 3.8679,
+      "step": 2028544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.677838102731788e-05,
+      "loss": 3.8779,
+      "step": 2029056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.676999507980736e-05,
+      "loss": 3.8586,
+      "step": 2029568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6761625511100572e-05,
+      "loss": 3.8773,
+      "step": 2030080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6753239563590052e-05,
+      "loss": 3.8644,
+      "step": 2030592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6744853616079532e-05,
+      "loss": 3.8643,
+      "step": 2031104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6736467668569012e-05,
+      "loss": 3.8626,
+      "step": 2031616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6728081721058492e-05,
+      "loss": 3.8725,
+      "step": 2032128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6719695773547972e-05,
+      "loss": 3.8745,
+      "step": 2032640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.671130982603745e-05,
+      "loss": 3.875,
+      "step": 2033152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.670292387852693e-05,
+      "loss": 3.858,
+      "step": 2033664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6694554309820144e-05,
+      "loss": 3.868,
+      "step": 2034176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.668616836230962e-05,
+      "loss": 3.8603,
+      "step": 2034688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.66777824147991e-05,
+      "loss": 3.8696,
+      "step": 2035200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.666939646728858e-05,
+      "loss": 3.8678,
+      "step": 2035712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6661026898581793e-05,
+      "loss": 3.8633,
+      "step": 2036224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6652657329875006e-05,
+      "loss": 3.8653,
+      "step": 2036736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6644271382364486e-05,
+      "loss": 3.8539,
+      "step": 2037248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6635885434853966e-05,
+      "loss": 3.8653,
+      "step": 2037760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6627499487343446e-05,
+      "loss": 3.8531,
+      "step": 2038272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6619129918636655e-05,
+      "loss": 3.8647,
+      "step": 2038784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6610743971126135e-05,
+      "loss": 3.8588,
+      "step": 2039296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6602358023615615e-05,
+      "loss": 3.8709,
+      "step": 2039808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6593972076105095e-05,
+      "loss": 3.8565,
+      "step": 2040320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6585586128594575e-05,
+      "loss": 3.8606,
+      "step": 2040832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6577200181084054e-05,
+      "loss": 3.8594,
+      "step": 2041344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6568814233573534e-05,
+      "loss": 3.874,
+      "step": 2041856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6560428286063014e-05,
+      "loss": 3.8668,
+      "step": 2042368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6552042338552494e-05,
+      "loss": 3.8673,
+      "step": 2042880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6543656391041974e-05,
+      "loss": 3.8692,
+      "step": 2043392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6535270443531454e-05,
+      "loss": 3.8765,
+      "step": 2043904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6526900874824667e-05,
+      "loss": 3.8646,
+      "step": 2044416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6518514927314147e-05,
+      "loss": 3.8648,
+      "step": 2044928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6510128979803627e-05,
+      "loss": 3.8645,
+      "step": 2045440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6501743032293103e-05,
+      "loss": 3.8716,
+      "step": 2045952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6493357084782583e-05,
+      "loss": 3.8652,
+      "step": 2046464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6484971137272066e-05,
+      "loss": 3.863,
+      "step": 2046976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6476601568565276e-05,
+      "loss": 3.8642,
+      "step": 2047488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6468215621054756e-05,
+      "loss": 3.8722,
+      "step": 2048000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6459829673544236e-05,
+      "loss": 3.8571,
+      "step": 2048512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6451443726033715e-05,
+      "loss": 3.8582,
+      "step": 2049024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6443057778523195e-05,
+      "loss": 3.8524,
+      "step": 2049536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6434671831012675e-05,
+      "loss": 3.8575,
+      "step": 2050048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.642628588350216e-05,
+      "loss": 3.8727,
+      "step": 2050560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6417899935991635e-05,
+      "loss": 3.8693,
+      "step": 2051072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6409530367284848e-05,
+      "loss": 3.8814,
+      "step": 2051584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6401144419774328e-05,
+      "loss": 3.8534,
+      "step": 2052096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6392758472263808e-05,
+      "loss": 3.8715,
+      "step": 2052608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6384372524753284e-05,
+      "loss": 3.867,
+      "step": 2053120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.63760029560465e-05,
+      "loss": 3.8702,
+      "step": 2053632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.636763338733971e-05,
+      "loss": 3.8637,
+      "step": 2054144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.635924743982919e-05,
+      "loss": 3.8625,
+      "step": 2054656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.635086149231867e-05,
+      "loss": 3.8666,
+      "step": 2055168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.634247554480815e-05,
+      "loss": 3.8811,
+      "step": 2055680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.633408959729763e-05,
+      "loss": 3.8531,
+      "step": 2056192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.632572002859084e-05,
+      "loss": 3.8689,
+      "step": 2056704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.631733408108032e-05,
+      "loss": 3.8693,
+      "step": 2057216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.63089481335698e-05,
+      "loss": 3.8663,
+      "step": 2057728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6300562186059278e-05,
+      "loss": 3.862,
+      "step": 2058240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6292176238548758e-05,
+      "loss": 3.8699,
+      "step": 2058752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6283823048645703e-05,
+      "loss": 3.8672,
+      "step": 2059264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6275437101135183e-05,
+      "loss": 3.8697,
+      "step": 2059776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6267051153624663e-05,
+      "loss": 3.8708,
+      "step": 2060288
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.000640869140625,
+      "eval_runtime": 302.5881,
+      "eval_samples_per_second": 1261.09,
+      "eval_steps_per_second": 39.41,
+      "step": 2060640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6258665206114143e-05,
+      "loss": 3.8497,
+      "step": 2060800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6250279258603623e-05,
+      "loss": 3.8606,
+      "step": 2061312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6241893311093103e-05,
+      "loss": 3.867,
+      "step": 2061824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6233507363582583e-05,
+      "loss": 3.8695,
+      "step": 2062336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6225121416072063e-05,
+      "loss": 3.8798,
+      "step": 2062848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6216735468561543e-05,
+      "loss": 3.865,
+      "step": 2063360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6208349521051023e-05,
+      "loss": 3.8724,
+      "step": 2063872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6199963573540503e-05,
+      "loss": 3.8648,
+      "step": 2064384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6191577626029982e-05,
+      "loss": 3.8607,
+      "step": 2064896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.618319167851946e-05,
+      "loss": 3.862,
+      "step": 2065408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.617480573100894e-05,
+      "loss": 3.8691,
+      "step": 2065920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6166419783498422e-05,
+      "loss": 3.8675,
+      "step": 2066432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6158050214791635e-05,
+      "loss": 3.8771,
+      "step": 2066944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.614966426728111e-05,
+      "loss": 3.8567,
+      "step": 2067456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.614127831977059e-05,
+      "loss": 3.8596,
+      "step": 2067968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.613289237226007e-05,
+      "loss": 3.8482,
+      "step": 2068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.612450642474955e-05,
+      "loss": 3.87,
+      "step": 2068992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.611612047723903e-05,
+      "loss": 3.868,
+      "step": 2069504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6107750908532244e-05,
+      "loss": 3.8548,
+      "step": 2070016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6099364961021724e-05,
+      "loss": 3.8707,
+      "step": 2070528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6090979013511204e-05,
+      "loss": 3.8865,
+      "step": 2071040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6082593066000684e-05,
+      "loss": 3.8713,
+      "step": 2071552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6074207118490164e-05,
+      "loss": 3.8696,
+      "step": 2072064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6065821170979643e-05,
+      "loss": 3.8653,
+      "step": 2072576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6057435223469123e-05,
+      "loss": 3.8718,
+      "step": 2073088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6049049275958603e-05,
+      "loss": 3.869,
+      "step": 2073600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6040663328448083e-05,
+      "loss": 3.8668,
+      "step": 2074112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6032277380937563e-05,
+      "loss": 3.8667,
+      "step": 2074624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6023891433427043e-05,
+      "loss": 3.8654,
+      "step": 2075136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6015521864720252e-05,
+      "loss": 3.8613,
+      "step": 2075648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6007135917209736e-05,
+      "loss": 3.8665,
+      "step": 2076160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5998749969699216e-05,
+      "loss": 3.8641,
+      "step": 2076672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5990364022188696e-05,
+      "loss": 3.869,
+      "step": 2077184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5981994453481905e-05,
+      "loss": 3.876,
+      "step": 2077696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5973608505971385e-05,
+      "loss": 3.8668,
+      "step": 2078208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5965238937264597e-05,
+      "loss": 3.866,
+      "step": 2078720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5956852989754077e-05,
+      "loss": 3.872,
+      "step": 2079232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5948467042243557e-05,
+      "loss": 3.8553,
+      "step": 2079744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5940081094733037e-05,
+      "loss": 3.8611,
+      "step": 2080256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5931695147222517e-05,
+      "loss": 3.8589,
+      "step": 2080768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5923309199711997e-05,
+      "loss": 3.8601,
+      "step": 2081280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5914923252201473e-05,
+      "loss": 3.8656,
+      "step": 2081792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5906537304690953e-05,
+      "loss": 3.8668,
+      "step": 2082304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.589816773598417e-05,
+      "loss": 3.8696,
+      "step": 2082816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.588978178847365e-05,
+      "loss": 3.8651,
+      "step": 2083328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5881395840963126e-05,
+      "loss": 3.8663,
+      "step": 2083840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5873009893452606e-05,
+      "loss": 3.8696,
+      "step": 2084352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5864623945942086e-05,
+      "loss": 3.8763,
+      "step": 2084864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5856237998431566e-05,
+      "loss": 3.8459,
+      "step": 2085376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5847868429724778e-05,
+      "loss": 3.8766,
+      "step": 2085888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5839482482214258e-05,
+      "loss": 3.8581,
+      "step": 2086400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5831096534703738e-05,
+      "loss": 3.8718,
+      "step": 2086912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5822710587193218e-05,
+      "loss": 3.8647,
+      "step": 2087424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5814324639682698e-05,
+      "loss": 3.8572,
+      "step": 2087936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5805938692172178e-05,
+      "loss": 3.8579,
+      "step": 2088448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.579756912346539e-05,
+      "loss": 3.8536,
+      "step": 2088960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.578918317595487e-05,
+      "loss": 3.8818,
+      "step": 2089472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.578079722844435e-05,
+      "loss": 3.8516,
+      "step": 2089984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.577241128093383e-05,
+      "loss": 3.8742,
+      "step": 2090496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.576404171222704e-05,
+      "loss": 3.8603,
+      "step": 2091008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.575565576471652e-05,
+      "loss": 3.8563,
+      "step": 2091520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5747269817206e-05,
+      "loss": 3.8718,
+      "step": 2092032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.573888386969548e-05,
+      "loss": 3.8573,
+      "step": 2092544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.573049792218496e-05,
+      "loss": 3.8447,
+      "step": 2093056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.572211197467444e-05,
+      "loss": 3.8538,
+      "step": 2093568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.571372602716392e-05,
+      "loss": 3.8745,
+      "step": 2094080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.57053400796534e-05,
+      "loss": 3.8565,
+      "step": 2094592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5696970510946608e-05,
+      "loss": 3.8717,
+      "step": 2095104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.568858456343609e-05,
+      "loss": 3.8466,
+      "step": 2095616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.568019861592557e-05,
+      "loss": 3.857,
+      "step": 2096128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.567182904721878e-05,
+      "loss": 3.8659,
+      "step": 2096640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.566344309970826e-05,
+      "loss": 3.8666,
+      "step": 2097152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.565505715219774e-05,
+      "loss": 3.8676,
+      "step": 2097664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.564667120468722e-05,
+      "loss": 3.8681,
+      "step": 2098176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5638301635980433e-05,
+      "loss": 3.873,
+      "step": 2098688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5629915688469913e-05,
+      "loss": 3.8592,
+      "step": 2099200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5621529740959393e-05,
+      "loss": 3.8638,
+      "step": 2099712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5613143793448873e-05,
+      "loss": 3.858,
+      "step": 2100224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5604757845938353e-05,
+      "loss": 3.8658,
+      "step": 2100736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5596371898427833e-05,
+      "loss": 3.8572,
+      "step": 2101248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.558798595091731e-05,
+      "loss": 3.8567,
+      "step": 2101760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5579600003406793e-05,
+      "loss": 3.8713,
+      "step": 2102272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5571230434700005e-05,
+      "loss": 3.8717,
+      "step": 2102784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5562844487189482e-05,
+      "loss": 3.8632,
+      "step": 2103296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.555445853967896e-05,
+      "loss": 3.843,
+      "step": 2103808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5546088970972174e-05,
+      "loss": 3.8663,
+      "step": 2104320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5537703023461654e-05,
+      "loss": 3.865,
+      "step": 2104832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5529317075951134e-05,
+      "loss": 3.8733,
+      "step": 2105344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5520931128440614e-05,
+      "loss": 3.8563,
+      "step": 2105856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5512561559733826e-05,
+      "loss": 3.8666,
+      "step": 2106368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5504175612223306e-05,
+      "loss": 3.8689,
+      "step": 2106880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5495789664712783e-05,
+      "loss": 3.857,
+      "step": 2107392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5487403717202263e-05,
+      "loss": 3.8563,
+      "step": 2107904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5479017769691746e-05,
+      "loss": 3.8677,
+      "step": 2108416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5470648200984955e-05,
+      "loss": 3.8646,
+      "step": 2108928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5462262253474435e-05,
+      "loss": 3.8722,
+      "step": 2109440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5453876305963915e-05,
+      "loss": 3.8579,
+      "step": 2109952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5445490358453395e-05,
+      "loss": 3.8617,
+      "step": 2110464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5437104410942875e-05,
+      "loss": 3.854,
+      "step": 2110976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5428734842236088e-05,
+      "loss": 3.8684,
+      "step": 2111488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5420348894725568e-05,
+      "loss": 3.8641,
+      "step": 2112000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5411962947215048e-05,
+      "loss": 3.8553,
+      "step": 2112512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5403576999704528e-05,
+      "loss": 3.8622,
+      "step": 2113024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5395191052194008e-05,
+      "loss": 3.854,
+      "step": 2113536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5386805104683487e-05,
+      "loss": 3.8605,
+      "step": 2114048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5378419157172967e-05,
+      "loss": 3.845,
+      "step": 2114560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.537004958846618e-05,
+      "loss": 3.8621,
+      "step": 2115072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.536166364095566e-05,
+      "loss": 3.8534,
+      "step": 2115584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5353277693445136e-05,
+      "loss": 3.8641,
+      "step": 2116096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5344891745934616e-05,
+      "loss": 3.8556,
+      "step": 2116608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.533652217722783e-05,
+      "loss": 3.8624,
+      "step": 2117120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.532813622971731e-05,
+      "loss": 3.852,
+      "step": 2117632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.531976666101052e-05,
+      "loss": 3.8693,
+      "step": 2118144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.53113807135e-05,
+      "loss": 3.8591,
+      "step": 2118656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.530299476598948e-05,
+      "loss": 3.861,
+      "step": 2119168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.529460881847896e-05,
+      "loss": 3.8628,
+      "step": 2119680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5286222870968438e-05,
+      "loss": 3.8764,
+      "step": 2120192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5277853302261654e-05,
+      "loss": 3.8595,
+      "step": 2120704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5269467354751134e-05,
+      "loss": 3.8595,
+      "step": 2121216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.526108140724061e-05,
+      "loss": 3.8624,
+      "step": 2121728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.525269545973009e-05,
+      "loss": 3.8666,
+      "step": 2122240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.524430951221957e-05,
+      "loss": 3.862,
+      "step": 2122752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5235923564709052e-05,
+      "loss": 3.8627,
+      "step": 2123264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5227537617198532e-05,
+      "loss": 3.8577,
+      "step": 2123776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5219151669688012e-05,
+      "loss": 3.8662,
+      "step": 2124288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5210765722177492e-05,
+      "loss": 3.8509,
+      "step": 2124800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.520237977466697e-05,
+      "loss": 3.8539,
+      "step": 2125312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.519399382715645e-05,
+      "loss": 3.8476,
+      "step": 2125824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5185624258449662e-05,
+      "loss": 3.8575,
+      "step": 2126336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5177238310939142e-05,
+      "loss": 3.8665,
+      "step": 2126848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.516885236342862e-05,
+      "loss": 3.8629,
+      "step": 2127360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.51604664159181e-05,
+      "loss": 3.8748,
+      "step": 2127872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5152096847211315e-05,
+      "loss": 3.8541,
+      "step": 2128384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5143710899700791e-05,
+      "loss": 3.863,
+      "step": 2128896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5135324952190271e-05,
+      "loss": 3.8685,
+      "step": 2129408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5126939004679753e-05,
+      "loss": 3.864,
+      "step": 2129920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5118569435972965e-05,
+      "loss": 3.8604,
+      "step": 2130432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5110183488462442e-05,
+      "loss": 3.8538,
+      "step": 2130944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5101797540951924e-05,
+      "loss": 3.8647,
+      "step": 2131456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5093411593441403e-05,
+      "loss": 3.8703,
+      "step": 2131968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5085025645930883e-05,
+      "loss": 3.856,
+      "step": 2132480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5076639698420363e-05,
+      "loss": 3.8642,
+      "step": 2132992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5068253750909845e-05,
+      "loss": 3.867,
+      "step": 2133504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5059867803399322e-05,
+      "loss": 3.8638,
+      "step": 2134016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5051481855888802e-05,
+      "loss": 3.8585,
+      "step": 2134528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5043128665985745e-05,
+      "loss": 3.8642,
+      "step": 2135040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5034742718475225e-05,
+      "loss": 3.8619,
+      "step": 2135552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5026356770964706e-05,
+      "loss": 3.8632,
+      "step": 2136064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5017987202257916e-05,
+      "loss": 3.8708,
+      "step": 2136576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.999410390853882,
+      "eval_runtime": 292.8768,
+      "eval_samples_per_second": 1302.906,
+      "eval_steps_per_second": 40.717,
+      "step": 2136960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5009601254747396e-05,
+      "loss": 3.8565,
+      "step": 2137088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5001215307236877e-05,
+      "loss": 3.8514,
+      "step": 2137600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4992829359726357e-05,
+      "loss": 3.8625,
+      "step": 2138112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4984443412215837e-05,
+      "loss": 3.8677,
+      "step": 2138624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4976057464705317e-05,
+      "loss": 3.8696,
+      "step": 2139136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4967671517194795e-05,
+      "loss": 3.8638,
+      "step": 2139648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4959285569684275e-05,
+      "loss": 3.8647,
+      "step": 2140160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4950899622173755e-05,
+      "loss": 3.8619,
+      "step": 2140672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4942513674663237e-05,
+      "loss": 3.8562,
+      "step": 2141184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4934144105956446e-05,
+      "loss": 3.8605,
+      "step": 2141696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4925758158445926e-05,
+      "loss": 3.8654,
+      "step": 2142208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4917372210935408e-05,
+      "loss": 3.8639,
+      "step": 2142720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.490900264222862e-05,
+      "loss": 3.8749,
+      "step": 2143232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4900616694718098e-05,
+      "loss": 3.8495,
+      "step": 2143744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4892230747207578e-05,
+      "loss": 3.8559,
+      "step": 2144256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4883844799697058e-05,
+      "loss": 3.8467,
+      "step": 2144768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4875475230990269e-05,
+      "loss": 3.8658,
+      "step": 2145280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4867089283479749e-05,
+      "loss": 3.8634,
+      "step": 2145792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4858703335969229e-05,
+      "loss": 3.8579,
+      "step": 2146304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4850317388458709e-05,
+      "loss": 3.8613,
+      "step": 2146816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4841931440948189e-05,
+      "loss": 3.8799,
+      "step": 2147328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.483354549343767e-05,
+      "loss": 3.8683,
+      "step": 2147840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.482515954592715e-05,
+      "loss": 3.8658,
+      "step": 2148352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4816773598416627e-05,
+      "loss": 3.8592,
+      "step": 2148864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4808387650906109e-05,
+      "loss": 3.8729,
+      "step": 2149376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4800001703395589e-05,
+      "loss": 3.862,
+      "step": 2149888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4791615755885069e-05,
+      "loss": 3.8588,
+      "step": 2150400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4783229808374549e-05,
+      "loss": 3.8681,
+      "step": 2150912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.477484386086403e-05,
+      "loss": 3.8563,
+      "step": 2151424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.476647429215724e-05,
+      "loss": 3.8594,
+      "step": 2151936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.475808834464672e-05,
+      "loss": 3.8638,
+      "step": 2152448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4749702397136201e-05,
+      "loss": 3.8592,
+      "step": 2152960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4741316449625681e-05,
+      "loss": 3.8671,
+      "step": 2153472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4732946880918892e-05,
+      "loss": 3.8743,
+      "step": 2153984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.47245773122121e-05,
+      "loss": 3.8589,
+      "step": 2154496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4716207743505315e-05,
+      "loss": 3.8635,
+      "step": 2155008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4707821795994795e-05,
+      "loss": 3.8688,
+      "step": 2155520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4699435848484275e-05,
+      "loss": 3.8514,
+      "step": 2156032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4691049900973753e-05,
+      "loss": 3.8552,
+      "step": 2156544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4682663953463233e-05,
+      "loss": 3.8585,
+      "step": 2157056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4674278005952713e-05,
+      "loss": 3.853,
+      "step": 2157568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4665892058442193e-05,
+      "loss": 3.8605,
+      "step": 2158080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4657506110931673e-05,
+      "loss": 3.8632,
+      "step": 2158592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4649120163421155e-05,
+      "loss": 3.8631,
+      "step": 2159104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4640734215910631e-05,
+      "loss": 3.8624,
+      "step": 2159616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4632348268400111e-05,
+      "loss": 3.8614,
+      "step": 2160128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4623978699693325e-05,
+      "loss": 3.8676,
+      "step": 2160640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4615592752182805e-05,
+      "loss": 3.8686,
+      "step": 2161152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4607206804672283e-05,
+      "loss": 3.8463,
+      "step": 2161664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4598820857161763e-05,
+      "loss": 3.8701,
+      "step": 2162176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4590434909651243e-05,
+      "loss": 3.853,
+      "step": 2162688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4582048962140723e-05,
+      "loss": 3.8638,
+      "step": 2163200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4573663014630203e-05,
+      "loss": 3.8647,
+      "step": 2163712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4565277067119685e-05,
+      "loss": 3.8538,
+      "step": 2164224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4556907498412894e-05,
+      "loss": 3.8532,
+      "step": 2164736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4548521550902374e-05,
+      "loss": 3.8485,
+      "step": 2165248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4540135603391856e-05,
+      "loss": 3.8798,
+      "step": 2165760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4531749655881336e-05,
+      "loss": 3.8491,
+      "step": 2166272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4523363708370812e-05,
+      "loss": 3.8659,
+      "step": 2166784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4514977760860294e-05,
+      "loss": 3.8596,
+      "step": 2167296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4506624570957235e-05,
+      "loss": 3.8537,
+      "step": 2167808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4498238623446717e-05,
+      "loss": 3.8626,
+      "step": 2168320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4489852675936197e-05,
+      "loss": 3.8578,
+      "step": 2168832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4481466728425677e-05,
+      "loss": 3.844,
+      "step": 2169344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4473080780915157e-05,
+      "loss": 3.8448,
+      "step": 2169856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4464694833404635e-05,
+      "loss": 3.8756,
+      "step": 2170368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4456308885894115e-05,
+      "loss": 3.8557,
+      "step": 2170880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4447939317187328e-05,
+      "loss": 3.8604,
+      "step": 2171392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.443955336967681e-05,
+      "loss": 3.8449,
+      "step": 2171904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4431167422166286e-05,
+      "loss": 3.8539,
+      "step": 2172416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4422781474655766e-05,
+      "loss": 3.8563,
+      "step": 2172928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4414395527145247e-05,
+      "loss": 3.8661,
+      "step": 2173440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4406009579634727e-05,
+      "loss": 3.8608,
+      "step": 2173952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4397623632124207e-05,
+      "loss": 3.8637,
+      "step": 2174464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4389237684613687e-05,
+      "loss": 3.8663,
+      "step": 2174976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.438088449471063e-05,
+      "loss": 3.8563,
+      "step": 2175488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4372498547200109e-05,
+      "loss": 3.8589,
+      "step": 2176000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4364112599689589e-05,
+      "loss": 3.8534,
+      "step": 2176512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4355726652179069e-05,
+      "loss": 3.8601,
+      "step": 2177024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4347340704668549e-05,
+      "loss": 3.8573,
+      "step": 2177536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4338954757158029e-05,
+      "loss": 3.8498,
+      "step": 2178048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.433058518845124e-05,
+      "loss": 3.8676,
+      "step": 2178560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.432219924094072e-05,
+      "loss": 3.8696,
+      "step": 2179072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4313813293430201e-05,
+      "loss": 3.86,
+      "step": 2179584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4305427345919681e-05,
+      "loss": 3.8369,
+      "step": 2180096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4297041398409161e-05,
+      "loss": 3.86,
+      "step": 2180608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4288655450898641e-05,
+      "loss": 3.861,
+      "step": 2181120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.428026950338812e-05,
+      "loss": 3.8681,
+      "step": 2181632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.42718835558776e-05,
+      "loss": 3.8579,
+      "step": 2182144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.426349760836708e-05,
+      "loss": 3.8617,
+      "step": 2182656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.425512803966029e-05,
+      "loss": 3.865,
+      "step": 2183168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.424674209214977e-05,
+      "loss": 3.8528,
+      "step": 2183680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.423835614463925e-05,
+      "loss": 3.8527,
+      "step": 2184192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4229970197128732e-05,
+      "loss": 3.8639,
+      "step": 2184704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.422160062842194e-05,
+      "loss": 3.8572,
+      "step": 2185216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4213231059715155e-05,
+      "loss": 3.8685,
+      "step": 2185728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4204845112204635e-05,
+      "loss": 3.8556,
+      "step": 2186240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4196459164694115e-05,
+      "loss": 3.8561,
+      "step": 2186752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4188073217183593e-05,
+      "loss": 3.8515,
+      "step": 2187264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4179703648476805e-05,
+      "loss": 3.8615,
+      "step": 2187776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4171317700966285e-05,
+      "loss": 3.8625,
+      "step": 2188288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4162931753455764e-05,
+      "loss": 3.8528,
+      "step": 2188800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4154545805945244e-05,
+      "loss": 3.8567,
+      "step": 2189312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4146159858434724e-05,
+      "loss": 3.8479,
+      "step": 2189824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4137773910924204e-05,
+      "loss": 3.8602,
+      "step": 2190336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4129387963413685e-05,
+      "loss": 3.8378,
+      "step": 2190848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4121002015903165e-05,
+      "loss": 3.8562,
+      "step": 2191360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4112648826000108e-05,
+      "loss": 3.8551,
+      "step": 2191872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4104262878489588e-05,
+      "loss": 3.8596,
+      "step": 2192384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4095876930979065e-05,
+      "loss": 3.847,
+      "step": 2192896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4087490983468547e-05,
+      "loss": 3.8602,
+      "step": 2193408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4079105035958027e-05,
+      "loss": 3.8489,
+      "step": 2193920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4070719088447507e-05,
+      "loss": 3.8628,
+      "step": 2194432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4062349519740717e-05,
+      "loss": 3.8557,
+      "step": 2194944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4053963572230197e-05,
+      "loss": 3.8573,
+      "step": 2195456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4045577624719677e-05,
+      "loss": 3.8567,
+      "step": 2195968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4037191677209157e-05,
+      "loss": 3.8736,
+      "step": 2196480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4028822108502368e-05,
+      "loss": 3.854,
+      "step": 2196992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4020436160991848e-05,
+      "loss": 3.8539,
+      "step": 2197504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4012050213481328e-05,
+      "loss": 3.8609,
+      "step": 2198016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.400366426597081e-05,
+      "loss": 3.8591,
+      "step": 2198528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.399527831846029e-05,
+      "loss": 3.8574,
+      "step": 2199040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.398689237094977e-05,
+      "loss": 3.8596,
+      "step": 2199552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3978506423439248e-05,
+      "loss": 3.8547,
+      "step": 2200064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.397013685473246e-05,
+      "loss": 3.8624,
+      "step": 2200576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.396175090722194e-05,
+      "loss": 3.8461,
+      "step": 2201088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3953364959711418e-05,
+      "loss": 3.8556,
+      "step": 2201600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3944979012200898e-05,
+      "loss": 3.8451,
+      "step": 2202112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3936609443494111e-05,
+      "loss": 3.8561,
+      "step": 2202624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3928223495983591e-05,
+      "loss": 3.8544,
+      "step": 2203136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3919837548473069e-05,
+      "loss": 3.8631,
+      "step": 2203648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3911451600962549e-05,
+      "loss": 3.8695,
+      "step": 2204160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3903065653452029e-05,
+      "loss": 3.8509,
+      "step": 2204672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.389467970594151e-05,
+      "loss": 3.8607,
+      "step": 2205184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.388629375843099e-05,
+      "loss": 3.8596,
+      "step": 2205696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.387790781092047e-05,
+      "loss": 3.8602,
+      "step": 2206208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3869554621017414e-05,
+      "loss": 3.8612,
+      "step": 2206720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.386116867350689e-05,
+      "loss": 3.8486,
+      "step": 2207232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3852782725996372e-05,
+      "loss": 3.8634,
+      "step": 2207744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3844396778485852e-05,
+      "loss": 3.8666,
+      "step": 2208256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3836010830975332e-05,
+      "loss": 3.8521,
+      "step": 2208768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3827624883464812e-05,
+      "loss": 3.8599,
+      "step": 2209280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3819238935954294e-05,
+      "loss": 3.8626,
+      "step": 2209792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3810852988443774e-05,
+      "loss": 3.8539,
+      "step": 2210304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.380246704093325e-05,
+      "loss": 3.8552,
+      "step": 2210816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3794097472226464e-05,
+      "loss": 3.8607,
+      "step": 2211328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3785711524715944e-05,
+      "loss": 3.8567,
+      "step": 2211840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3777325577205421e-05,
+      "loss": 3.8573,
+      "step": 2212352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3768939629694903e-05,
+      "loss": 3.8682,
+      "step": 2212864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.997990846633911,
+      "eval_runtime": 294.2748,
+      "eval_samples_per_second": 1296.717,
+      "eval_steps_per_second": 40.523,
+      "step": 2213280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3760570060988115e-05,
+      "loss": 3.8628,
+      "step": 2213376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3752184113477595e-05,
+      "loss": 3.8461,
+      "step": 2213888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3743798165967073e-05,
+      "loss": 3.8584,
+      "step": 2214400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3735412218456553e-05,
+      "loss": 3.8662,
+      "step": 2214912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3727026270946033e-05,
+      "loss": 3.8628,
+      "step": 2215424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3718640323435513e-05,
+      "loss": 3.8615,
+      "step": 2215936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3710254375924995e-05,
+      "loss": 3.8563,
+      "step": 2216448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3701868428414475e-05,
+      "loss": 3.8586,
+      "step": 2216960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3693482480903955e-05,
+      "loss": 3.852,
+      "step": 2217472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3685096533393433e-05,
+      "loss": 3.8585,
+      "step": 2217984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3676710585882913e-05,
+      "loss": 3.8582,
+      "step": 2218496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3668324638372393e-05,
+      "loss": 3.8603,
+      "step": 2219008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3659955069665604e-05,
+      "loss": 3.871,
+      "step": 2219520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3651569122155084e-05,
+      "loss": 3.8462,
+      "step": 2220032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3643183174644564e-05,
+      "loss": 3.8498,
+      "step": 2220544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3634797227134043e-05,
+      "loss": 3.8416,
+      "step": 2221056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3626411279623525e-05,
+      "loss": 3.8605,
+      "step": 2221568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3618025332113005e-05,
+      "loss": 3.8583,
+      "step": 2222080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3609655763406214e-05,
+      "loss": 3.8557,
+      "step": 2222592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3601269815895696e-05,
+      "loss": 3.8565,
+      "step": 2223104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3592883868385176e-05,
+      "loss": 3.8746,
+      "step": 2223616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3584497920874656e-05,
+      "loss": 3.8635,
+      "step": 2224128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3576111973364136e-05,
+      "loss": 3.8629,
+      "step": 2224640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3567726025853614e-05,
+      "loss": 3.8552,
+      "step": 2225152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3559340078343094e-05,
+      "loss": 3.8735,
+      "step": 2225664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3550954130832574e-05,
+      "loss": 3.8549,
+      "step": 2226176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3542568183322056e-05,
+      "loss": 3.8578,
+      "step": 2226688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3534182235811535e-05,
+      "loss": 3.862,
+      "step": 2227200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3525796288301015e-05,
+      "loss": 3.8523,
+      "step": 2227712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3517410340790492e-05,
+      "loss": 3.8562,
+      "step": 2228224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3509057150887435e-05,
+      "loss": 3.8589,
+      "step": 2228736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3500671203376917e-05,
+      "loss": 3.8524,
+      "step": 2229248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3492285255866397e-05,
+      "loss": 3.8639,
+      "step": 2229760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3483899308355877e-05,
+      "loss": 3.8691,
+      "step": 2230272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3475513360845357e-05,
+      "loss": 3.8568,
+      "step": 2230784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3467143792138568e-05,
+      "loss": 3.8589,
+      "step": 2231296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3458757844628048e-05,
+      "loss": 3.8696,
+      "step": 2231808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3450371897117528e-05,
+      "loss": 3.8401,
+      "step": 2232320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3441985949607007e-05,
+      "loss": 3.8551,
+      "step": 2232832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3433600002096489e-05,
+      "loss": 3.8549,
+      "step": 2233344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3425214054585966e-05,
+      "loss": 3.8467,
+      "step": 2233856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3416828107075446e-05,
+      "loss": 3.8591,
+      "step": 2234368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3408442159564927e-05,
+      "loss": 3.8631,
+      "step": 2234880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.340007259085814e-05,
+      "loss": 3.8589,
+      "step": 2235392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.339170302215135e-05,
+      "loss": 3.8582,
+      "step": 2235904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.338331707464083e-05,
+      "loss": 3.8548,
+      "step": 2236416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.337493112713031e-05,
+      "loss": 3.8601,
+      "step": 2236928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.336654517961979e-05,
+      "loss": 3.8648,
+      "step": 2237440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3358159232109269e-05,
+      "loss": 3.8458,
+      "step": 2237952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3349773284598749e-05,
+      "loss": 3.8634,
+      "step": 2238464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3341387337088229e-05,
+      "loss": 3.8553,
+      "step": 2238976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.333300138957771e-05,
+      "loss": 3.8576,
+      "step": 2239488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.332463182087092e-05,
+      "loss": 3.8614,
+      "step": 2240000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.33162458733604e-05,
+      "loss": 3.8454,
+      "step": 2240512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3307859925849881e-05,
+      "loss": 3.8555,
+      "step": 2241024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.329949035714309e-05,
+      "loss": 3.8428,
+      "step": 2241536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3291104409632572e-05,
+      "loss": 3.8752,
+      "step": 2242048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3282718462122052e-05,
+      "loss": 3.844,
+      "step": 2242560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3274332514611532e-05,
+      "loss": 3.8661,
+      "step": 2243072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3265946567101012e-05,
+      "loss": 3.851,
+      "step": 2243584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3257576998394222e-05,
+      "loss": 3.8512,
+      "step": 2244096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3249191050883702e-05,
+      "loss": 3.8574,
+      "step": 2244608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3240805103373182e-05,
+      "loss": 3.8576,
+      "step": 2245120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3232435534666393e-05,
+      "loss": 3.8439,
+      "step": 2245632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3224049587155873e-05,
+      "loss": 3.8365,
+      "step": 2246144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3215663639645353e-05,
+      "loss": 3.8667,
+      "step": 2246656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3207277692134835e-05,
+      "loss": 3.8567,
+      "step": 2247168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3198891744624315e-05,
+      "loss": 3.8556,
+      "step": 2247680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3190505797113795e-05,
+      "loss": 3.8422,
+      "step": 2248192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3182119849603273e-05,
+      "loss": 3.8485,
+      "step": 2248704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3173750280896485e-05,
+      "loss": 3.8551,
+      "step": 2249216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3165364333385965e-05,
+      "loss": 3.8607,
+      "step": 2249728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3156978385875444e-05,
+      "loss": 3.8582,
+      "step": 2250240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3148592438364923e-05,
+      "loss": 3.861,
+      "step": 2250752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3140206490854403e-05,
+      "loss": 3.8632,
+      "step": 2251264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3131836922147616e-05,
+      "loss": 3.8507,
+      "step": 2251776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3123467353440827e-05,
+      "loss": 3.8558,
+      "step": 2252288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3115081405930307e-05,
+      "loss": 3.85,
+      "step": 2252800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3106695458419788e-05,
+      "loss": 3.8567,
+      "step": 2253312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3098309510909268e-05,
+      "loss": 3.8541,
+      "step": 2253824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3089923563398745e-05,
+      "loss": 3.8443,
+      "step": 2254336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3081537615888226e-05,
+      "loss": 3.8612,
+      "step": 2254848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3073151668377706e-05,
+      "loss": 3.8657,
+      "step": 2255360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3064765720867186e-05,
+      "loss": 3.8545,
+      "step": 2255872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3056379773356666e-05,
+      "loss": 3.8386,
+      "step": 2256384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3048010204649877e-05,
+      "loss": 3.8563,
+      "step": 2256896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3039624257139357e-05,
+      "loss": 3.8601,
+      "step": 2257408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3031238309628837e-05,
+      "loss": 3.8606,
+      "step": 2257920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3022852362118319e-05,
+      "loss": 3.8503,
+      "step": 2258432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3014482793411528e-05,
+      "loss": 3.8569,
+      "step": 2258944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3006096845901008e-05,
+      "loss": 3.863,
+      "step": 2259456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.299771089839049e-05,
+      "loss": 3.8472,
+      "step": 2259968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.298932495087997e-05,
+      "loss": 3.8512,
+      "step": 2260480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.298093900336945e-05,
+      "loss": 3.8556,
+      "step": 2260992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2972553055858928e-05,
+      "loss": 3.8554,
+      "step": 2261504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.296418348715214e-05,
+      "loss": 3.8668,
+      "step": 2262016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.295579753964162e-05,
+      "loss": 3.8572,
+      "step": 2262528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2947411592131098e-05,
+      "loss": 3.8468,
+      "step": 2263040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2939025644620578e-05,
+      "loss": 3.8487,
+      "step": 2263552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2930639697110058e-05,
+      "loss": 3.8573,
+      "step": 2264064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2922253749599538e-05,
+      "loss": 3.8597,
+      "step": 2264576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.291386780208902e-05,
+      "loss": 3.8501,
+      "step": 2265088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.29054818545785e-05,
+      "loss": 3.8507,
+      "step": 2265600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.289709590706798e-05,
+      "loss": 3.8423,
+      "step": 2266112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.288872633836119e-05,
+      "loss": 3.8591,
+      "step": 2266624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.288034039085067e-05,
+      "loss": 3.8344,
+      "step": 2267136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2871987200947614e-05,
+      "loss": 3.8503,
+      "step": 2267648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2863601253437094e-05,
+      "loss": 3.8515,
+      "step": 2268160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2855215305926572e-05,
+      "loss": 3.8567,
+      "step": 2268672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2846829358416052e-05,
+      "loss": 3.8454,
+      "step": 2269184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2838443410905532e-05,
+      "loss": 3.8548,
+      "step": 2269696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2830057463395012e-05,
+      "loss": 3.8451,
+      "step": 2270208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2821671515884492e-05,
+      "loss": 3.8582,
+      "step": 2270720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2813285568373973e-05,
+      "loss": 3.8546,
+      "step": 2271232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2804899620863453e-05,
+      "loss": 3.8517,
+      "step": 2271744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.279651367335293e-05,
+      "loss": 3.854,
+      "step": 2272256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2788127725842412e-05,
+      "loss": 3.8673,
+      "step": 2272768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2779758157135624e-05,
+      "loss": 3.8527,
+      "step": 2273280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2771372209625104e-05,
+      "loss": 3.8535,
+      "step": 2273792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2762986262114582e-05,
+      "loss": 3.8532,
+      "step": 2274304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2754600314604062e-05,
+      "loss": 3.8529,
+      "step": 2274816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2746230745897275e-05,
+      "loss": 3.8538,
+      "step": 2275328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2737844798386753e-05,
+      "loss": 3.855,
+      "step": 2275840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2729458850876233e-05,
+      "loss": 3.8522,
+      "step": 2276352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2721072903365713e-05,
+      "loss": 3.8578,
+      "step": 2276864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2712686955855193e-05,
+      "loss": 3.8452,
+      "step": 2277376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2704317387148404e-05,
+      "loss": 3.8487,
+      "step": 2277888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2695931439637884e-05,
+      "loss": 3.8444,
+      "step": 2278400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2687545492127365e-05,
+      "loss": 3.8512,
+      "step": 2278912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2679159544616845e-05,
+      "loss": 3.8445,
+      "step": 2279424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2670773597106325e-05,
+      "loss": 3.8636,
+      "step": 2279936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2662387649595805e-05,
+      "loss": 3.8636,
+      "step": 2280448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2654001702085283e-05,
+      "loss": 3.8454,
+      "step": 2280960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2645615754574763e-05,
+      "loss": 3.8567,
+      "step": 2281472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2637246185867976e-05,
+      "loss": 3.8574,
+      "step": 2281984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2628876617161187e-05,
+      "loss": 3.8601,
+      "step": 2282496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2620490669650667e-05,
+      "loss": 3.8492,
+      "step": 2283008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2612104722140147e-05,
+      "loss": 3.8473,
+      "step": 2283520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2603718774629628e-05,
+      "loss": 3.8624,
+      "step": 2284032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2595332827119108e-05,
+      "loss": 3.8613,
+      "step": 2284544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2586946879608585e-05,
+      "loss": 3.8467,
+      "step": 2285056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2578560932098066e-05,
+      "loss": 3.8565,
+      "step": 2285568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2570174984587546e-05,
+      "loss": 3.8623,
+      "step": 2286080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2561805415880757e-05,
+      "loss": 3.8509,
+      "step": 2286592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2553419468370237e-05,
+      "loss": 3.8478,
+      "step": 2287104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2545033520859717e-05,
+      "loss": 3.8579,
+      "step": 2287616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2536647573349197e-05,
+      "loss": 3.8503,
+      "step": 2288128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2528278004642408e-05,
+      "loss": 3.8586,
+      "step": 2288640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2519892057131888e-05,
+      "loss": 3.86,
+      "step": 2289152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.996950626373291,
+      "eval_runtime": 295.69,
+      "eval_samples_per_second": 1290.51,
+      "eval_steps_per_second": 40.329,
+      "step": 2289600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2511506109621368e-05,
+      "loss": 3.8499,
+      "step": 2289664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2503120162110848e-05,
+      "loss": 3.8423,
+      "step": 2290176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.249473421460033e-05,
+      "loss": 3.8535,
+      "step": 2290688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2486348267089808e-05,
+      "loss": 3.8656,
+      "step": 2291200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2477962319579288e-05,
+      "loss": 3.8592,
+      "step": 2291712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2469576372068767e-05,
+      "loss": 3.8572,
+      "step": 2292224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2461190424558247e-05,
+      "loss": 3.8516,
+      "step": 2292736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2452804477047727e-05,
+      "loss": 3.8554,
+      "step": 2293248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2444418529537207e-05,
+      "loss": 3.8472,
+      "step": 2293760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2436032582026689e-05,
+      "loss": 3.8574,
+      "step": 2294272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.242767939212363e-05,
+      "loss": 3.8507,
+      "step": 2294784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.241929344461311e-05,
+      "loss": 3.8607,
+      "step": 2295296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2410923875906321e-05,
+      "loss": 3.8648,
+      "step": 2295808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2402537928395801e-05,
+      "loss": 3.8448,
+      "step": 2296320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2394151980885281e-05,
+      "loss": 3.8452,
+      "step": 2296832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2385766033374761e-05,
+      "loss": 3.8369,
+      "step": 2297344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2377380085864241e-05,
+      "loss": 3.8567,
+      "step": 2297856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2368994138353721e-05,
+      "loss": 3.8578,
+      "step": 2298368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2360624569646932e-05,
+      "loss": 3.8525,
+      "step": 2298880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2352238622136412e-05,
+      "loss": 3.8501,
+      "step": 2299392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2343852674625894e-05,
+      "loss": 3.875,
+      "step": 2299904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2335466727115372e-05,
+      "loss": 3.8617,
+      "step": 2300416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2327080779604852e-05,
+      "loss": 3.8559,
+      "step": 2300928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2318694832094332e-05,
+      "loss": 3.855,
+      "step": 2301440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2310308884583812e-05,
+      "loss": 3.8676,
+      "step": 2301952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2301922937073292e-05,
+      "loss": 3.8519,
+      "step": 2302464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2293553368366504e-05,
+      "loss": 3.8509,
+      "step": 2302976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2285167420855984e-05,
+      "loss": 3.8632,
+      "step": 2303488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2276797852149193e-05,
+      "loss": 3.8473,
+      "step": 2304000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2268411904638675e-05,
+      "loss": 3.8501,
+      "step": 2304512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2260025957128155e-05,
+      "loss": 3.8527,
+      "step": 2305024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2251640009617635e-05,
+      "loss": 3.8484,
+      "step": 2305536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2243254062107113e-05,
+      "loss": 3.859,
+      "step": 2306048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2234868114596595e-05,
+      "loss": 3.8651,
+      "step": 2306560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2226482167086075e-05,
+      "loss": 3.8519,
+      "step": 2307072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2218112598379285e-05,
+      "loss": 3.8626,
+      "step": 2307584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2209726650868765e-05,
+      "loss": 3.8598,
+      "step": 2308096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2201340703358245e-05,
+      "loss": 3.8362,
+      "step": 2308608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2192954755847725e-05,
+      "loss": 3.85,
+      "step": 2309120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2184568808337205e-05,
+      "loss": 3.8547,
+      "step": 2309632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2176182860826685e-05,
+      "loss": 3.84,
+      "step": 2310144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2167813292119896e-05,
+      "loss": 3.8583,
+      "step": 2310656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2159427344609376e-05,
+      "loss": 3.8605,
+      "step": 2311168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2151041397098856e-05,
+      "loss": 3.8524,
+      "step": 2311680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2142655449588336e-05,
+      "loss": 3.8549,
+      "step": 2312192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2134269502077814e-05,
+      "loss": 3.8566,
+      "step": 2312704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2125883554567296e-05,
+      "loss": 3.851,
+      "step": 2313216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2117497607056776e-05,
+      "loss": 3.8648,
+      "step": 2313728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2109111659546256e-05,
+      "loss": 3.8395,
+      "step": 2314240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2100725712035736e-05,
+      "loss": 3.8584,
+      "step": 2314752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2092339764525216e-05,
+      "loss": 3.8563,
+      "step": 2315264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2083953817014696e-05,
+      "loss": 3.8523,
+      "step": 2315776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2075567869504174e-05,
+      "loss": 3.8569,
+      "step": 2316288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2067198300797386e-05,
+      "loss": 3.8394,
+      "step": 2316800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2058812353286866e-05,
+      "loss": 3.8537,
+      "step": 2317312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2050426405776346e-05,
+      "loss": 3.8362,
+      "step": 2317824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2042056837069557e-05,
+      "loss": 3.8702,
+      "step": 2318336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2033670889559037e-05,
+      "loss": 3.8397,
+      "step": 2318848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2025284942048517e-05,
+      "loss": 3.8601,
+      "step": 2319360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2016898994537997e-05,
+      "loss": 3.8525,
+      "step": 2319872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2008513047027477e-05,
+      "loss": 3.8472,
+      "step": 2320384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2000143478320688e-05,
+      "loss": 3.8494,
+      "step": 2320896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.199175753081017e-05,
+      "loss": 3.8547,
+      "step": 2321408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1983371583299647e-05,
+      "loss": 3.8362,
+      "step": 2321920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1974985635789127e-05,
+      "loss": 3.8355,
+      "step": 2322432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1966599688278609e-05,
+      "loss": 3.8618,
+      "step": 2322944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1958213740768087e-05,
+      "loss": 3.8524,
+      "step": 2323456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1949827793257567e-05,
+      "loss": 3.8497,
+      "step": 2323968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1941441845747047e-05,
+      "loss": 3.8373,
+      "step": 2324480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.193307227704026e-05,
+      "loss": 3.8461,
+      "step": 2324992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.192470270833347e-05,
+      "loss": 3.8528,
+      "step": 2325504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.191631676082295e-05,
+      "loss": 3.8568,
+      "step": 2326016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.190793081331243e-05,
+      "loss": 3.8563,
+      "step": 2326528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.189954486580191e-05,
+      "loss": 3.8565,
+      "step": 2327040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.189115891829139e-05,
+      "loss": 3.8562,
+      "step": 2327552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.188277297078087e-05,
+      "loss": 3.8478,
+      "step": 2328064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.187438702327035e-05,
+      "loss": 3.8488,
+      "step": 2328576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1866001075759829e-05,
+      "loss": 3.8466,
+      "step": 2329088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1857647885856772e-05,
+      "loss": 3.855,
+      "step": 2329600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1849261938346252e-05,
+      "loss": 3.8505,
+      "step": 2330112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1840892369639464e-05,
+      "loss": 3.8432,
+      "step": 2330624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1832506422128943e-05,
+      "loss": 3.8566,
+      "step": 2331136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1824120474618424e-05,
+      "loss": 3.8589,
+      "step": 2331648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1815734527107904e-05,
+      "loss": 3.8522,
+      "step": 2332160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1807348579597384e-05,
+      "loss": 3.838,
+      "step": 2332672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1798962632086862e-05,
+      "loss": 3.8469,
+      "step": 2333184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1790576684576344e-05,
+      "loss": 3.853,
+      "step": 2333696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1782207115869555e-05,
+      "loss": 3.861,
+      "step": 2334208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1773821168359033e-05,
+      "loss": 3.8508,
+      "step": 2334720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1765435220848515e-05,
+      "loss": 3.8514,
+      "step": 2335232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1757049273337995e-05,
+      "loss": 3.8567,
+      "step": 2335744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1748663325827475e-05,
+      "loss": 3.8477,
+      "step": 2336256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1740277378316953e-05,
+      "loss": 3.8441,
+      "step": 2336768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1731891430806435e-05,
+      "loss": 3.8501,
+      "step": 2337280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1723505483295915e-05,
+      "loss": 3.8543,
+      "step": 2337792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1715119535785393e-05,
+      "loss": 3.859,
+      "step": 2338304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1706733588274873e-05,
+      "loss": 3.8573,
+      "step": 2338816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1698347640764354e-05,
+      "loss": 3.8402,
+      "step": 2339328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1689961693253833e-05,
+      "loss": 3.8457,
+      "step": 2339840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1681592124547045e-05,
+      "loss": 3.8511,
+      "step": 2340352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1673206177036525e-05,
+      "loss": 3.8571,
+      "step": 2340864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1664836608329736e-05,
+      "loss": 3.8484,
+      "step": 2341376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1656450660819216e-05,
+      "loss": 3.8478,
+      "step": 2341888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1648064713308696e-05,
+      "loss": 3.8414,
+      "step": 2342400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1639678765798176e-05,
+      "loss": 3.853,
+      "step": 2342912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1631292818287656e-05,
+      "loss": 3.8325,
+      "step": 2343424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1622923249580866e-05,
+      "loss": 3.8469,
+      "step": 2343936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1614553680874079e-05,
+      "loss": 3.8506,
+      "step": 2344448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1606167733363559e-05,
+      "loss": 3.8494,
+      "step": 2344960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1597781785853039e-05,
+      "loss": 3.8441,
+      "step": 2345472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1589395838342517e-05,
+      "loss": 3.8507,
+      "step": 2345984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1581009890831999e-05,
+      "loss": 3.8431,
+      "step": 2346496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1572623943321479e-05,
+      "loss": 3.8505,
+      "step": 2347008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1564237995810957e-05,
+      "loss": 3.852,
+      "step": 2347520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1555852048300437e-05,
+      "loss": 3.8453,
+      "step": 2348032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1547466100789919e-05,
+      "loss": 3.8554,
+      "step": 2348544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1539080153279397e-05,
+      "loss": 3.861,
+      "step": 2349056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1530710584572608e-05,
+      "loss": 3.855,
+      "step": 2349568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.152232463706209e-05,
+      "loss": 3.8414,
+      "step": 2350080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.151393868955157e-05,
+      "loss": 3.8551,
+      "step": 2350592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1505552742041048e-05,
+      "loss": 3.8501,
+      "step": 2351104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.149718317333426e-05,
+      "loss": 3.8458,
+      "step": 2351616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.148879722582374e-05,
+      "loss": 3.8532,
+      "step": 2352128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1480411278313218e-05,
+      "loss": 3.8488,
+      "step": 2352640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.14720253308027e-05,
+      "loss": 3.8549,
+      "step": 2353152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.146363938329218e-05,
+      "loss": 3.8438,
+      "step": 2353664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.145525343578166e-05,
+      "loss": 3.846,
+      "step": 2354176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1446867488271138e-05,
+      "loss": 3.8384,
+      "step": 2354688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.143849791956435e-05,
+      "loss": 3.847,
+      "step": 2355200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.143011197205383e-05,
+      "loss": 3.8383,
+      "step": 2355712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.142172602454331e-05,
+      "loss": 3.8614,
+      "step": 2356224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.141334007703279e-05,
+      "loss": 3.8636,
+      "step": 2356736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.140495412952227e-05,
+      "loss": 3.8463,
+      "step": 2357248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.139656818201175e-05,
+      "loss": 3.8486,
+      "step": 2357760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1388198613304961e-05,
+      "loss": 3.858,
+      "step": 2358272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1379812665794441e-05,
+      "loss": 3.8529,
+      "step": 2358784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1371426718283921e-05,
+      "loss": 3.8463,
+      "step": 2359296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1363040770773401e-05,
+      "loss": 3.8443,
+      "step": 2359808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1354654823262881e-05,
+      "loss": 3.8579,
+      "step": 2360320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1346268875752361e-05,
+      "loss": 3.8572,
+      "step": 2360832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1337899307045573e-05,
+      "loss": 3.8447,
+      "step": 2361344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1329513359535052e-05,
+      "loss": 3.8516,
+      "step": 2361856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1321127412024532e-05,
+      "loss": 3.8579,
+      "step": 2362368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1312741464514012e-05,
+      "loss": 3.8451,
+      "step": 2362880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1304355517003491e-05,
+      "loss": 3.8477,
+      "step": 2363392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1295969569492971e-05,
+      "loss": 3.8568,
+      "step": 2363904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1287583621982451e-05,
+      "loss": 3.8474,
+      "step": 2364416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1279197674471931e-05,
+      "loss": 3.8552,
+      "step": 2364928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1270828105765142e-05,
+      "loss": 3.855,
+      "step": 2365440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.995748996734619,
+      "eval_runtime": 296.0059,
+      "eval_samples_per_second": 1289.133,
+      "eval_steps_per_second": 40.286,
+      "step": 2365920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1262442158254622e-05,
+      "loss": 3.8354,
+      "step": 2365952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1254056210744104e-05,
+      "loss": 3.8414,
+      "step": 2366464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1245670263233582e-05,
+      "loss": 3.8483,
+      "step": 2366976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1237284315723062e-05,
+      "loss": 3.8603,
+      "step": 2367488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1228898368212542e-05,
+      "loss": 3.8543,
+      "step": 2368000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1220512420702024e-05,
+      "loss": 3.8588,
+      "step": 2368512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1212126473191502e-05,
+      "loss": 3.8457,
+      "step": 2369024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1203740525680982e-05,
+      "loss": 3.8557,
+      "step": 2369536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1195354578170462e-05,
+      "loss": 3.8408,
+      "step": 2370048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1186968630659942e-05,
+      "loss": 3.856,
+      "step": 2370560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1178582683149422e-05,
+      "loss": 3.8461,
+      "step": 2371072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1170196735638902e-05,
+      "loss": 3.8581,
+      "step": 2371584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1161827166932114e-05,
+      "loss": 3.8624,
+      "step": 2372096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1153441219421592e-05,
+      "loss": 3.8413,
+      "step": 2372608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1145055271911072e-05,
+      "loss": 3.8429,
+      "step": 2373120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1136669324400554e-05,
+      "loss": 3.8358,
+      "step": 2373632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1128283376890032e-05,
+      "loss": 3.8454,
+      "step": 2374144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1119897429379512e-05,
+      "loss": 3.8522,
+      "step": 2374656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1111511481868992e-05,
+      "loss": 3.8504,
+      "step": 2375168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1103125534358472e-05,
+      "loss": 3.8442,
+      "step": 2375680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1094739586847952e-05,
+      "loss": 3.8748,
+      "step": 2376192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1086353639337432e-05,
+      "loss": 3.8555,
+      "step": 2376704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1077967691826912e-05,
+      "loss": 3.8551,
+      "step": 2377216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1069581744316392e-05,
+      "loss": 3.8496,
+      "step": 2377728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1061228554413335e-05,
+      "loss": 3.8643,
+      "step": 2378240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1052842606902815e-05,
+      "loss": 3.8478,
+      "step": 2378752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1044456659392295e-05,
+      "loss": 3.8494,
+      "step": 2379264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1036070711881773e-05,
+      "loss": 3.8574,
+      "step": 2379776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1027684764371255e-05,
+      "loss": 3.8415,
+      "step": 2380288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1019298816860735e-05,
+      "loss": 3.8513,
+      "step": 2380800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1010912869350213e-05,
+      "loss": 3.8469,
+      "step": 2381312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1002543300643426e-05,
+      "loss": 3.8477,
+      "step": 2381824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0994157353132906e-05,
+      "loss": 3.8551,
+      "step": 2382336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0985771405622384e-05,
+      "loss": 3.8615,
+      "step": 2382848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0977385458111866e-05,
+      "loss": 3.8492,
+      "step": 2383360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0969015889405076e-05,
+      "loss": 3.8616,
+      "step": 2383872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0960629941894556e-05,
+      "loss": 3.8553,
+      "step": 2384384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0952243994384036e-05,
+      "loss": 3.8331,
+      "step": 2384896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0943858046873516e-05,
+      "loss": 3.8498,
+      "step": 2385408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0935472099362996e-05,
+      "loss": 3.8487,
+      "step": 2385920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0927086151852475e-05,
+      "loss": 3.8332,
+      "step": 2386432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0918700204341956e-05,
+      "loss": 3.8561,
+      "step": 2386944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0910314256831436e-05,
+      "loss": 3.8556,
+      "step": 2387456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0901928309320916e-05,
+      "loss": 3.8514,
+      "step": 2387968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0893542361810394e-05,
+      "loss": 3.8441,
+      "step": 2388480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0885156414299876e-05,
+      "loss": 3.8569,
+      "step": 2388992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0876770466789356e-05,
+      "loss": 3.8475,
+      "step": 2389504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0868400898082567e-05,
+      "loss": 3.8622,
+      "step": 2390016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0860014950572047e-05,
+      "loss": 3.8365,
+      "step": 2390528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0851645381865257e-05,
+      "loss": 3.8509,
+      "step": 2391040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0843259434354739e-05,
+      "loss": 3.8563,
+      "step": 2391552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0834873486844217e-05,
+      "loss": 3.8477,
+      "step": 2392064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0826487539333697e-05,
+      "loss": 3.8589,
+      "step": 2392576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0818101591823177e-05,
+      "loss": 3.8349,
+      "step": 2393088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0809715644312657e-05,
+      "loss": 3.8524,
+      "step": 2393600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0801329696802137e-05,
+      "loss": 3.8354,
+      "step": 2394112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0792960128095348e-05,
+      "loss": 3.8632,
+      "step": 2394624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.078457418058483e-05,
+      "loss": 3.8405,
+      "step": 2395136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0776188233074308e-05,
+      "loss": 3.8537,
+      "step": 2395648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0767802285563788e-05,
+      "loss": 3.8513,
+      "step": 2396160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0759416338053268e-05,
+      "loss": 3.8436,
+      "step": 2396672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0751030390542748e-05,
+      "loss": 3.8452,
+      "step": 2397184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0742644443032228e-05,
+      "loss": 3.8521,
+      "step": 2397696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.073427487432544e-05,
+      "loss": 3.833,
+      "step": 2398208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.072588892681492e-05,
+      "loss": 3.8334,
+      "step": 2398720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0717502979304398e-05,
+      "loss": 3.8555,
+      "step": 2399232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0709117031793878e-05,
+      "loss": 3.8505,
+      "step": 2399744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.070073108428336e-05,
+      "loss": 3.8489,
+      "step": 2400256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0692345136772838e-05,
+      "loss": 3.835,
+      "step": 2400768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0683959189262318e-05,
+      "loss": 3.842,
+      "step": 2401280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.067558962055553e-05,
+      "loss": 3.8477,
+      "step": 2401792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.066720367304501e-05,
+      "loss": 3.8511,
+      "step": 2402304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0658817725534489e-05,
+      "loss": 3.8529,
+      "step": 2402816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0650448156827701e-05,
+      "loss": 3.8509,
+      "step": 2403328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0642062209317181e-05,
+      "loss": 3.8554,
+      "step": 2403840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0633676261806661e-05,
+      "loss": 3.848,
+      "step": 2404352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0625290314296141e-05,
+      "loss": 3.8437,
+      "step": 2404864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0616904366785621e-05,
+      "loss": 3.8471,
+      "step": 2405376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0608518419275101e-05,
+      "loss": 3.8461,
+      "step": 2405888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0600148850568312e-05,
+      "loss": 3.8489,
+      "step": 2406400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0591762903057792e-05,
+      "loss": 3.8424,
+      "step": 2406912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0583376955547272e-05,
+      "loss": 3.8465,
+      "step": 2407424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0574991008036752e-05,
+      "loss": 3.8624,
+      "step": 2407936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0566621439329963e-05,
+      "loss": 3.845,
+      "step": 2408448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0558235491819443e-05,
+      "loss": 3.837,
+      "step": 2408960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0549849544308924e-05,
+      "loss": 3.8433,
+      "step": 2409472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0541463596798403e-05,
+      "loss": 3.8492,
+      "step": 2409984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0533094028091613e-05,
+      "loss": 3.8542,
+      "step": 2410496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0524708080581095e-05,
+      "loss": 3.8481,
+      "step": 2411008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0516322133070575e-05,
+      "loss": 3.8451,
+      "step": 2411520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0507936185560053e-05,
+      "loss": 3.856,
+      "step": 2412032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0499550238049533e-05,
+      "loss": 3.8473,
+      "step": 2412544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0491164290539015e-05,
+      "loss": 3.841,
+      "step": 2413056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0482778343028493e-05,
+      "loss": 3.8445,
+      "step": 2413568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0474392395517973e-05,
+      "loss": 3.8502,
+      "step": 2414080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0466006448007453e-05,
+      "loss": 3.8566,
+      "step": 2414592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0457636879300665e-05,
+      "loss": 3.8524,
+      "step": 2415104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0449250931790144e-05,
+      "loss": 3.8402,
+      "step": 2415616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0440881363083356e-05,
+      "loss": 3.8425,
+      "step": 2416128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0432495415572836e-05,
+      "loss": 3.8473,
+      "step": 2416640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0424109468062316e-05,
+      "loss": 3.8574,
+      "step": 2417152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0415723520551796e-05,
+      "loss": 3.8409,
+      "step": 2417664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0407337573041276e-05,
+      "loss": 3.8419,
+      "step": 2418176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0398951625530756e-05,
+      "loss": 3.8408,
+      "step": 2418688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0390565678020236e-05,
+      "loss": 3.8457,
+      "step": 2419200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0382179730509716e-05,
+      "loss": 3.831,
+      "step": 2419712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0373793782999196e-05,
+      "loss": 3.8441,
+      "step": 2420224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0365424214292407e-05,
+      "loss": 3.8424,
+      "step": 2420736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0357038266781887e-05,
+      "loss": 3.8473,
+      "step": 2421248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0348652319271367e-05,
+      "loss": 3.8413,
+      "step": 2421760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0340266371760847e-05,
+      "loss": 3.8486,
+      "step": 2422272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0331880424250326e-05,
+      "loss": 3.8371,
+      "step": 2422784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0323510855543537e-05,
+      "loss": 3.8462,
+      "step": 2423296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0315124908033017e-05,
+      "loss": 3.8506,
+      "step": 2423808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0306738960522497e-05,
+      "loss": 3.8394,
+      "step": 2424320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0298353013011977e-05,
+      "loss": 3.8516,
+      "step": 2424832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0289967065501457e-05,
+      "loss": 3.8581,
+      "step": 2425344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.028159749679467e-05,
+      "loss": 3.852,
+      "step": 2425856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0273211549284148e-05,
+      "loss": 3.8402,
+      "step": 2426368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0264825601773628e-05,
+      "loss": 3.8447,
+      "step": 2426880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.025643965426311e-05,
+      "loss": 3.8493,
+      "step": 2427392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0248053706752588e-05,
+      "loss": 3.8418,
+      "step": 2427904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0239667759242068e-05,
+      "loss": 3.8527,
+      "step": 2428416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0231281811731548e-05,
+      "loss": 3.8432,
+      "step": 2428928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.022289586422103e-05,
+      "loss": 3.8539,
+      "step": 2429440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0214526295514238e-05,
+      "loss": 3.8387,
+      "step": 2429952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0206140348003718e-05,
+      "loss": 3.844,
+      "step": 2430464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.01977544004932e-05,
+      "loss": 3.8344,
+      "step": 2430976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.018938483178641e-05,
+      "loss": 3.8417,
+      "step": 2431488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.018099888427589e-05,
+      "loss": 3.8321,
+      "step": 2432000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.017261293676537e-05,
+      "loss": 3.8549,
+      "step": 2432512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.016422698925485e-05,
+      "loss": 3.8633,
+      "step": 2433024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0155841041744329e-05,
+      "loss": 3.8434,
+      "step": 2433536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.014745509423381e-05,
+      "loss": 3.8393,
+      "step": 2434048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.013906914672329e-05,
+      "loss": 3.8553,
+      "step": 2434560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0130683199212769e-05,
+      "loss": 3.8535,
+      "step": 2435072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0122330009309712e-05,
+      "loss": 3.8451,
+      "step": 2435584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0113960440602925e-05,
+      "loss": 3.8394,
+      "step": 2436096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0105574493092404e-05,
+      "loss": 3.8551,
+      "step": 2436608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0097188545581884e-05,
+      "loss": 3.8487,
+      "step": 2437120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0088802598071363e-05,
+      "loss": 3.8445,
+      "step": 2437632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0080416650560844e-05,
+      "loss": 3.8452,
+      "step": 2438144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0072030703050324e-05,
+      "loss": 3.8529,
+      "step": 2438656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0063644755539803e-05,
+      "loss": 3.8449,
+      "step": 2439168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0055258808029283e-05,
+      "loss": 3.8425,
+      "step": 2439680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0046872860518764e-05,
+      "loss": 3.8526,
+      "step": 2440192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0038486913008242e-05,
+      "loss": 3.8442,
+      "step": 2440704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0030117344301453e-05,
+      "loss": 3.8483,
+      "step": 2441216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0021731396790935e-05,
+      "loss": 3.855,
+      "step": 2441728
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0013345449280415e-05,
+      "loss": 3.8443,
+      "step": 2442240
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9946417808532715,
+      "eval_runtime": 303.6538,
+      "eval_samples_per_second": 1256.665,
+      "eval_steps_per_second": 39.272,
+      "step": 2442240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0004959501769893e-05,
+      "loss": 3.838,
+      "step": 2442752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.996573554259373e-06,
+      "loss": 3.842,
+      "step": 2443264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.988187606748855e-06,
+      "loss": 3.8571,
+      "step": 2443776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.979818038042065e-06,
+      "loss": 3.8529,
+      "step": 2444288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.971432090531545e-06,
+      "loss": 3.856,
+      "step": 2444800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.963046143021025e-06,
+      "loss": 3.8407,
+      "step": 2445312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.954660195510505e-06,
+      "loss": 3.8534,
+      "step": 2445824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.946290626803716e-06,
+      "loss": 3.8361,
+      "step": 2446336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.937904679293196e-06,
+      "loss": 3.8544,
+      "step": 2446848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.929518731782676e-06,
+      "loss": 3.8384,
+      "step": 2447360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.921132784272156e-06,
+      "loss": 3.8586,
+      "step": 2447872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.912746836761636e-06,
+      "loss": 3.8568,
+      "step": 2448384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.90439364685858e-06,
+      "loss": 3.8419,
+      "step": 2448896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.89602407815179e-06,
+      "loss": 3.8382,
+      "step": 2449408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.88763813064127e-06,
+      "loss": 3.8366,
+      "step": 2449920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.87925218313075e-06,
+      "loss": 3.8367,
+      "step": 2450432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.87086623562023e-06,
+      "loss": 3.85,
+      "step": 2450944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.86248028810971e-06,
+      "loss": 3.8457,
+      "step": 2451456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.854094340599188e-06,
+      "loss": 3.844,
+      "step": 2451968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.84570839308867e-06,
+      "loss": 3.8693,
+      "step": 2452480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.83732244557815e-06,
+      "loss": 3.851,
+      "step": 2452992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.82893649806763e-06,
+      "loss": 3.8545,
+      "step": 2453504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.820550550557108e-06,
+      "loss": 3.8442,
+      "step": 2454016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.81216460304659e-06,
+      "loss": 3.8596,
+      "step": 2454528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.80377865553607e-06,
+      "loss": 3.8414,
+      "step": 2455040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.795392708025548e-06,
+      "loss": 3.8535,
+      "step": 2455552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.78702313931876e-06,
+      "loss": 3.8492,
+      "step": 2456064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.77863719180824e-06,
+      "loss": 3.8437,
+      "step": 2456576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.770251244297719e-06,
+      "loss": 3.8444,
+      "step": 2457088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.7618652967872e-06,
+      "loss": 3.8427,
+      "step": 2457600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.753495728080411e-06,
+      "loss": 3.8461,
+      "step": 2458112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.745126159373623e-06,
+      "loss": 3.8495,
+      "step": 2458624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.736740211863102e-06,
+      "loss": 3.8541,
+      "step": 2459136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.728354264352582e-06,
+      "loss": 3.8491,
+      "step": 2459648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.719984695645794e-06,
+      "loss": 3.8545,
+      "step": 2460160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.711598748135274e-06,
+      "loss": 3.8514,
+      "step": 2460672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.703212800624752e-06,
+      "loss": 3.8345,
+      "step": 2461184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.694826853114234e-06,
+      "loss": 3.845,
+      "step": 2461696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.686440905603714e-06,
+      "loss": 3.8417,
+      "step": 2462208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.678054958093192e-06,
+      "loss": 3.8334,
+      "step": 2462720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.669669010582672e-06,
+      "loss": 3.8493,
+      "step": 2463232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.661283063072154e-06,
+      "loss": 3.8554,
+      "step": 2463744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.652897115561634e-06,
+      "loss": 3.8446,
+      "step": 2464256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.644511168051112e-06,
+      "loss": 3.8439,
+      "step": 2464768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.636125220540592e-06,
+      "loss": 3.8516,
+      "step": 2465280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.627739273030074e-06,
+      "loss": 3.8498,
+      "step": 2465792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.619369704323283e-06,
+      "loss": 3.8541,
+      "step": 2466304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.610983756812764e-06,
+      "loss": 3.8376,
+      "step": 2466816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.602597809302244e-06,
+      "loss": 3.8429,
+      "step": 2467328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.594211861791724e-06,
+      "loss": 3.8506,
+      "step": 2467840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.585842293084935e-06,
+      "loss": 3.8461,
+      "step": 2468352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.577456345574415e-06,
+      "loss": 3.8485,
+      "step": 2468864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.569070398063895e-06,
+      "loss": 3.835,
+      "step": 2469376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.560684450553373e-06,
+      "loss": 3.8491,
+      "step": 2469888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.552298503042855e-06,
+      "loss": 3.8288,
+      "step": 2470400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.543928934336066e-06,
+      "loss": 3.8574,
+      "step": 2470912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.535542986825546e-06,
+      "loss": 3.8393,
+      "step": 2471424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.527157039315026e-06,
+      "loss": 3.8482,
+      "step": 2471936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.518771091804506e-06,
+      "loss": 3.8435,
+      "step": 2472448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.510385144293986e-06,
+      "loss": 3.8445,
+      "step": 2472960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.501999196783466e-06,
+      "loss": 3.8386,
+      "step": 2473472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.493613249272945e-06,
+      "loss": 3.8492,
+      "step": 2473984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.485227301762425e-06,
+      "loss": 3.8324,
+      "step": 2474496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.476857733055638e-06,
+      "loss": 3.8295,
+      "step": 2475008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.468488164348847e-06,
+      "loss": 3.8488,
+      "step": 2475520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.460102216838327e-06,
+      "loss": 3.852,
+      "step": 2476032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.451716269327809e-06,
+      "loss": 3.8439,
+      "step": 2476544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.44334670062102e-06,
+      "loss": 3.8304,
+      "step": 2477056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.4349607531105e-06,
+      "loss": 3.8386,
+      "step": 2477568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.42657480559998e-06,
+      "loss": 3.8449,
+      "step": 2478080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.41818885808946e-06,
+      "loss": 3.8487,
+      "step": 2478592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.409802910578938e-06,
+      "loss": 3.8454,
+      "step": 2479104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.40141696306842e-06,
+      "loss": 3.8479,
+      "step": 2479616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.3930310155579e-06,
+      "loss": 3.8526,
+      "step": 2480128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.384645068047379e-06,
+      "loss": 3.8502,
+      "step": 2480640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.376259120536857e-06,
+      "loss": 3.8334,
+      "step": 2481152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.367873173026339e-06,
+      "loss": 3.8425,
+      "step": 2481664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.359487225515819e-06,
+      "loss": 3.8494,
+      "step": 2482176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.351117656809028e-06,
+      "loss": 3.8423,
+      "step": 2482688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.34273170929851e-06,
+      "loss": 3.8392,
+      "step": 2483200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.33434576178799e-06,
+      "loss": 3.8412,
+      "step": 2483712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.32595981427747e-06,
+      "loss": 3.8597,
+      "step": 2484224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.31757386676695e-06,
+      "loss": 3.8408,
+      "step": 2484736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.30920429806016e-06,
+      "loss": 3.8311,
+      "step": 2485248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.30081835054964e-06,
+      "loss": 3.8419,
+      "step": 2485760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.292448781842853e-06,
+      "loss": 3.8453,
+      "step": 2486272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.284062834332331e-06,
+      "loss": 3.8529,
+      "step": 2486784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.275676886821811e-06,
+      "loss": 3.841,
+      "step": 2487296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.267290939311293e-06,
+      "loss": 3.8433,
+      "step": 2487808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.258904991800771e-06,
+      "loss": 3.8473,
+      "step": 2488320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.250519044290251e-06,
+      "loss": 3.8471,
+      "step": 2488832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.242149475583463e-06,
+      "loss": 3.8345,
+      "step": 2489344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.233763528072943e-06,
+      "loss": 3.8432,
+      "step": 2489856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.225377580562422e-06,
+      "loss": 3.8497,
+      "step": 2490368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.216991633051902e-06,
+      "loss": 3.8564,
+      "step": 2490880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.208605685541383e-06,
+      "loss": 3.8432,
+      "step": 2491392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.200219738030861e-06,
+      "loss": 3.8354,
+      "step": 2491904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.191833790520341e-06,
+      "loss": 3.8426,
+      "step": 2492416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.183447843009823e-06,
+      "loss": 3.8402,
+      "step": 2492928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.175061895499301e-06,
+      "loss": 3.8526,
+      "step": 2493440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.166675947988781e-06,
+      "loss": 3.8398,
+      "step": 2493952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.158290000478261e-06,
+      "loss": 3.8399,
+      "step": 2494464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.149920431771474e-06,
+      "loss": 3.8354,
+      "step": 2494976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.141534484260952e-06,
+      "loss": 3.847,
+      "step": 2495488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.133148536750432e-06,
+      "loss": 3.8226,
+      "step": 2496000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.124762589239914e-06,
+      "loss": 3.8426,
+      "step": 2496512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.116393020533123e-06,
+      "loss": 3.8348,
+      "step": 2497024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.108007073022604e-06,
+      "loss": 3.8472,
+      "step": 2497536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.099621125512084e-06,
+      "loss": 3.8357,
+      "step": 2498048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.091235178001564e-06,
+      "loss": 3.8452,
+      "step": 2498560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.082849230491043e-06,
+      "loss": 3.8358,
+      "step": 2499072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.074463282980524e-06,
+      "loss": 3.8421,
+      "step": 2499584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.066077335470004e-06,
+      "loss": 3.8454,
+      "step": 2500096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.057691387959482e-06,
+      "loss": 3.835,
+      "step": 2500608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.049305440448962e-06,
+      "loss": 3.8522,
+      "step": 2501120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.040919492938444e-06,
+      "loss": 3.8562,
+      "step": 2501632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.032549924231655e-06,
+      "loss": 3.8474,
+      "step": 2502144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.024163976721133e-06,
+      "loss": 3.8388,
+      "step": 2502656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.015778029210615e-06,
+      "loss": 3.8419,
+      "step": 2503168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.007392081700095e-06,
+      "loss": 3.8447,
+      "step": 2503680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.999006134189573e-06,
+      "loss": 3.8396,
+      "step": 2504192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.990636565482785e-06,
+      "loss": 3.8489,
+      "step": 2504704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.982266996775996e-06,
+      "loss": 3.8406,
+      "step": 2505216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.973881049265478e-06,
+      "loss": 3.8466,
+      "step": 2505728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.965495101754956e-06,
+      "loss": 3.8355,
+      "step": 2506240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.957109154244436e-06,
+      "loss": 3.8392,
+      "step": 2506752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.948723206733916e-06,
+      "loss": 3.8298,
+      "step": 2507264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.940337259223396e-06,
+      "loss": 3.8352,
+      "step": 2507776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.931967690516607e-06,
+      "loss": 3.8297,
+      "step": 2508288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.923581743006087e-06,
+      "loss": 3.8525,
+      "step": 2508800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.915195795495568e-06,
+      "loss": 3.8599,
+      "step": 2509312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.906809847985047e-06,
+      "loss": 3.8457,
+      "step": 2509824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.898423900474527e-06,
+      "loss": 3.83,
+      "step": 2510336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.890037952964007e-06,
+      "loss": 3.8525,
+      "step": 2510848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.881652005453486e-06,
+      "loss": 3.8486,
+      "step": 2511360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.873266057942966e-06,
+      "loss": 3.842,
+      "step": 2511872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.864880110432446e-06,
+      "loss": 3.8367,
+      "step": 2512384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.856494162921928e-06,
+      "loss": 3.8511,
+      "step": 2512896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.848124594215137e-06,
+      "loss": 3.848,
+      "step": 2513408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.839738646704617e-06,
+      "loss": 3.8415,
+      "step": 2513920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.831352699194099e-06,
+      "loss": 3.8434,
+      "step": 2514432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.822966751683577e-06,
+      "loss": 3.8466,
+      "step": 2514944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.814580804173057e-06,
+      "loss": 3.8384,
+      "step": 2515456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.806194856662537e-06,
+      "loss": 3.84,
+      "step": 2515968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.797808909152019e-06,
+      "loss": 3.8526,
+      "step": 2516480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.789422961641497e-06,
+      "loss": 3.8413,
+      "step": 2516992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.78105339293471e-06,
+      "loss": 3.8437,
+      "step": 2517504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.77268382422792e-06,
+      "loss": 3.8538,
+      "step": 2518016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.7642978767174e-06,
+      "loss": 3.8383,
+      "step": 2518528
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9933865070343018,
+      "eval_runtime": 301.046,
+      "eval_samples_per_second": 1267.551,
+      "eval_steps_per_second": 39.612,
+      "step": 2518560
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.0084651268735483e+18,
+  "trial_name": null,
+  "trial_params": null
+}