diff --git "a/checkpoint-915840/trainer_state.json" "b/checkpoint-915840/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-915840/trainer_state.json"
@@ -0,0 +1,10849 @@
+{
+  "best_metric": 4.063039302825928,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/lstm/1/checkpoints/checkpoint-915840",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 915840,
+  "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": 10.8219,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5524,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0539,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9881,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9415,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9135,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7447,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6404,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5312,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4514,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3765,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3144,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2444,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1729,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1148,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0464,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0047,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.963,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9229,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.8876,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.86,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.82,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.7966,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7535,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.742,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979036769104073e-05,
+      "loss": 5.7094,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.978198174353021e-05,
+      "loss": 5.6906,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6635,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6304,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975684027980238e-05,
+      "loss": 5.6147,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974845433229186e-05,
+      "loss": 5.5904,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 5.58,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5546,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.544,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5332,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5105,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.4994,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4566,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4533,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.4331,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.418,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4137,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 5.4059,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 5.378,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.3773,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3731,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3602,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9605991497435414e-05,
+      "loss": 5.3537,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 5.3371,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3156,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.319,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.2998,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.302,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2714,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2839,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.2566,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2741,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2457,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 5.2339,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 5.2289,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.2234,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2096,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2111,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 5.1837,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 5.1796,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.1922,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.1839,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1665,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.15,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1492,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1499,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1428,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1395,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1347,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1353,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1275,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1021,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.0971,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.0947,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0824,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.091,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0741,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.0916,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 5.0663,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 5.0513,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0543,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.0539,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.053,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0392,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0342,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0263,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 5.0248,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0143,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0101,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0083,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.014,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.9945,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9869,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.981,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9774,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.9816,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.9684,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.9614,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.953,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.9569,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9278,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9435,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9319,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9383,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9254,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9339,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9195,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9122,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.9206,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.8991,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9152,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.9048,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9022,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.8892,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.8899,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.8921,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8823,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.8817,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.8725,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.874,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8551,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8591,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.843,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8493,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8356,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8563,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8617,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8463,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8252,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8357,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8339,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8288,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8229,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8232,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8204,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8099,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.8104,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.81,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.7984,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.7985,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.8062,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.7932,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.7923,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.8054,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.7815,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.770177364349365,
+      "eval_runtime": 685.2663,
+      "eval_samples_per_second": 556.851,
+      "eval_steps_per_second": 17.402,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.7697,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7756,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7891,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7763,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7824,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7571,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7749,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7465,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7618,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667092952331785e-05,
+      "loss": 4.7497,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7648,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865032105731075e-05,
+      "loss": 4.759,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7469,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7415,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7288,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7295,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7397,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7359,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.7285,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583249856030314e-05,
+      "loss": 4.7433,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7366,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7287,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7172,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.7377,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7128,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7221,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7175,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7113,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7019,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.7016,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.7034,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.7007,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.7011,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846589572729423e-05,
+      "loss": 4.7063,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.6972,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.6938,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6752,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6813,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6717,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6686,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.6795,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.6807,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6634,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838208538860022e-05,
+      "loss": 4.6707,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6689,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.6719,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.6733,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.6653,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834015565104763e-05,
+      "loss": 4.6534,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833176970353711e-05,
+      "loss": 4.6613,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832338375602659e-05,
+      "loss": 4.65,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6585,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6451,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6559,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6344,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.6582,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.641,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.6366,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.644,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247942986039366e-05,
+      "loss": 4.6343,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239557038528846e-05,
+      "loss": 4.6275,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6423,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6109,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6179,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820606238489796e-05,
+      "loss": 4.6302,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819767643738744e-05,
+      "loss": 4.6296,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.6198,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6113,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.6112,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6163,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.621,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.6143,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81389748048138e-05,
+      "loss": 4.6169,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.6228,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812220290979276e-05,
+      "loss": 4.6195,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811383334108597e-05,
+      "loss": 4.6006,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810544739357545e-05,
+      "loss": 4.5996,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.6093,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808867549855441e-05,
+      "loss": 4.5929,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.6048,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.5902,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.6143,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.596,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804676213980554e-05,
+      "loss": 4.5812,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803837619229502e-05,
+      "loss": 4.5909,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.5923,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.5963,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.5867,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.5849,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.587,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.5792,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5776,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5774,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.5841,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795454947479728e-05,
+      "loss": 4.5828,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794616352728676e-05,
+      "loss": 4.5768,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793781033738371e-05,
+      "loss": 4.5631,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792942438987319e-05,
+      "loss": 4.566,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.571,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.5697,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5664,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895896978634836e-05,
+      "loss": 4.5612,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887511031124316e-05,
+      "loss": 4.5543,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879125083613796e-05,
+      "loss": 4.5659,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870739136103276e-05,
+      "loss": 4.535,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862353188592756e-05,
+      "loss": 4.5587,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.5502,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.557,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.5457,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.5583,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.5464,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.542,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803667934822844e-05,
+      "loss": 4.5551,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795281987312324e-05,
+      "loss": 4.5383,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786896039801804e-05,
+      "loss": 4.5514,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778510092291284e-05,
+      "loss": 4.5508,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777014052358449e-05,
+      "loss": 4.545,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776175457607397e-05,
+      "loss": 4.5357,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.5401,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5442,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.528,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.5461,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.5288,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5363,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.5274,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.5271,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5085,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.5229,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766954191106572e-05,
+      "loss": 4.5096,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76611559635552e-05,
+      "loss": 4.5337,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765277001604468e-05,
+      "loss": 4.544,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764438406853416e-05,
+      "loss": 4.5266,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763599812102364e-05,
+      "loss": 4.5107,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762761217351312e-05,
+      "loss": 4.5233,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76192262260026e-05,
+      "loss": 4.5176,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610856657295813e-05,
+      "loss": 4.5208,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760248708858902e-05,
+      "loss": 4.5182,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941011410785e-05,
+      "loss": 4.5167,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.5119,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757732924605746e-05,
+      "loss": 4.5115,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756894329854694e-05,
+      "loss": 4.5109,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756055735103642e-05,
+      "loss": 4.516,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.5061,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5013,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753541588730859e-05,
+      "loss": 4.5152,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752702993979807e-05,
+      "loss": 4.5097,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751864399228755e-05,
+      "loss": 4.5007,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751025804477703e-05,
+      "loss": 4.5181,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750187209726651e-05,
+      "loss": 4.4981,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.491509437561035,
+      "eval_runtime": 664.5011,
+      "eval_samples_per_second": 574.252,
+      "eval_steps_per_second": 17.946,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749348614975599e-05,
+      "loss": 4.4867,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748510020224547e-05,
+      "loss": 4.4917,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747671425473495e-05,
+      "loss": 4.5088,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746832830722443e-05,
+      "loss": 4.499,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745994235971391e-05,
+      "loss": 4.5036,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745155641220339e-05,
+      "loss": 4.4832,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744317046469287e-05,
+      "loss": 4.5011,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743478451718235e-05,
+      "loss": 4.4744,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742639856967183e-05,
+      "loss": 4.4976,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741801262216131e-05,
+      "loss": 4.4804,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740962667465079e-05,
+      "loss": 4.4987,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401240727140264e-05,
+      "loss": 4.4949,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739287115843348e-05,
+      "loss": 4.4837,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384485210922966e-05,
+      "loss": 4.4787,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376099263412446e-05,
+      "loss": 4.4731,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4727,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.48,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4802,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734255547337036e-05,
+      "loss": 4.4796,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733416952585984e-05,
+      "loss": 4.4896,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732578357834932e-05,
+      "loss": 4.4875,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73173976308388e-05,
+      "loss": 4.4818,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730901168332828e-05,
+      "loss": 4.4731,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4852,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4703,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4771,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.4757,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.4708,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4619,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4644,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.465,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.467,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4726,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721683177592748e-05,
+      "loss": 4.4676,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720844582841696e-05,
+      "loss": 4.4637,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720005988090644e-05,
+      "loss": 4.4637,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719167393339592e-05,
+      "loss": 4.4476,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71832879858854e-05,
+      "loss": 4.457,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717490203837488e-05,
+      "loss": 4.4485,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716651609086437e-05,
+      "loss": 4.4464,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715813014335385e-05,
+      "loss": 4.4518,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.4596,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714137462713654e-05,
+      "loss": 4.4455,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4474,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4558,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711621678460498e-05,
+      "loss": 4.4468,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4605,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.446,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709105894207341e-05,
+      "loss": 4.4343,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708267299456289e-05,
+      "loss": 4.4483,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074303425856106e-05,
+      "loss": 4.4391,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065917478345586e-05,
+      "loss": 4.4463,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057531530835066e-05,
+      "loss": 4.4346,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049145583324546e-05,
+      "loss": 4.4414,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040776014617755e-05,
+      "loss": 4.4316,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032390067107235e-05,
+      "loss": 4.4459,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024004119596715e-05,
+      "loss": 4.4347,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.434,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.44,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4282,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.4223,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982074382044115e-05,
+      "loss": 4.4447,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4114,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965318865826804e-05,
+      "loss": 4.4155,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695693291831629e-05,
+      "loss": 4.4255,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694854697080577e-05,
+      "loss": 4.4364,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4242,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693179145458846e-05,
+      "loss": 4.4145,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4181,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.418,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.43,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4231,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688987809583959e-05,
+      "loss": 4.4222,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4282,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4307,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4113,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856350684601245e-05,
+      "loss": 4.4106,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847964737090725e-05,
+      "loss": 4.4215,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839595168383934e-05,
+      "loss": 4.404,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831209220873414e-05,
+      "loss": 4.4195,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682283965216662e-05,
+      "loss": 4.4042,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.4266,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680606775714558e-05,
+      "loss": 4.4135,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.4008,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678929586212454e-05,
+      "loss": 4.4051,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.4108,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677254034590723e-05,
+      "loss": 4.4207,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676415439839671e-05,
+      "loss": 4.405,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675576845088619e-05,
+      "loss": 4.4091,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.4011,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.4105,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4002,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672222466084412e-05,
+      "loss": 4.3986,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671383871333359e-05,
+      "loss": 4.4102,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670545276582307e-05,
+      "loss": 4.407,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669706681831255e-05,
+      "loss": 4.4059,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668868087080203e-05,
+      "loss": 4.3918,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668029492329151e-05,
+      "loss": 4.3949,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667192535458472e-05,
+      "loss": 4.3977,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66635394070742e-05,
+      "loss": 4.3993,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665515345956368e-05,
+      "loss": 4.3966,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664676751205316e-05,
+      "loss": 4.3874,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.3877,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630011995835856e-05,
+      "loss": 4.397,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.3697,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.3891,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660488691091176e-05,
+      "loss": 4.3784,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596500963401234e-05,
+      "loss": 4.3946,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588115015890714e-05,
+      "loss": 4.3838,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579729068380194e-05,
+      "loss": 4.3922,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571343120869674e-05,
+      "loss": 4.3831,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562957173359154e-05,
+      "loss": 4.3789,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554571225848634e-05,
+      "loss": 4.3917,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546185278338114e-05,
+      "loss": 4.3795,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537832088435065e-05,
+      "loss": 4.3907,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652944614092454e-05,
+      "loss": 4.3875,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652106019341402e-05,
+      "loss": 4.3881,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65126742459035e-05,
+      "loss": 4.3776,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650428829839298e-05,
+      "loss": 4.3739,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649590235088246e-05,
+      "loss": 4.3887,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648753278217567e-05,
+      "loss": 4.37,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647914683466515e-05,
+      "loss": 4.3878,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647076088715463e-05,
+      "loss": 4.3722,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.379,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645398899213359e-05,
+      "loss": 4.3752,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644560304462307e-05,
+      "loss": 4.373,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437217097112554e-05,
+      "loss": 4.3552,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642884752840576e-05,
+      "loss": 4.3725,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.3601,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641209201218845e-05,
+      "loss": 4.3754,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640370606467793e-05,
+      "loss": 4.3876,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639532011716741e-05,
+      "loss": 4.3838,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638693416965689e-05,
+      "loss": 4.3527,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637854822214637e-05,
+      "loss": 4.3772,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637016227463585e-05,
+      "loss": 4.3721,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636177632712533e-05,
+      "loss": 4.3697,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635339037961481e-05,
+      "loss": 4.3608,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634500443210429e-05,
+      "loss": 4.3722,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633661848459377e-05,
+      "loss": 4.3648,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632823253708325e-05,
+      "loss": 4.3632,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631984658957273e-05,
+      "loss": 4.3652,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631146064206221e-05,
+      "loss": 4.3684,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630310745215916e-05,
+      "loss": 4.3629,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629472150464864e-05,
+      "loss": 4.3605,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628633555713812e-05,
+      "loss": 4.3662,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277949609627597e-05,
+      "loss": 4.3657,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626956366211707e-05,
+      "loss": 4.3571,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626117771460655e-05,
+      "loss": 4.3725,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625279176709603e-05,
+      "loss": 4.3568,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.357151508331299,
+      "eval_runtime": 289.0299,
+      "eval_samples_per_second": 1320.248,
+      "eval_steps_per_second": 41.259,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624440581958551e-05,
+      "loss": 4.3505,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.3498,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227650303368205e-05,
+      "loss": 4.3679,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219264355857685e-05,
+      "loss": 4.3548,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210878408347165e-05,
+      "loss": 4.3685,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.3433,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194106513326125e-05,
+      "loss": 4.363,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185720565815605e-05,
+      "loss": 4.3326,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177334618305085e-05,
+      "loss": 4.3636,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.345,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160579102087774e-05,
+      "loss": 4.3586,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152193154577254e-05,
+      "loss": 4.3569,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614382358587046e-05,
+      "loss": 4.3423,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613543763835994e-05,
+      "loss": 4.3482,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612705169084942e-05,
+      "loss": 4.3389,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611866574333891e-05,
+      "loss": 4.3358,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611027979582839e-05,
+      "loss": 4.3492,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.3447,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.347,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3544,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3596,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3489,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605996411076526e-05,
+      "loss": 4.3413,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605157816325474e-05,
+      "loss": 4.3533,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604320859454796e-05,
+      "loss": 4.3384,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603483902584117e-05,
+      "loss": 4.351,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602645307833065e-05,
+      "loss": 4.3458,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601806713082013e-05,
+      "loss": 4.3372,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600968118330961e-05,
+      "loss": 4.3316,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600129523579909e-05,
+      "loss": 4.3387,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599290928828857e-05,
+      "loss": 4.334,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598452334077805e-05,
+      "loss": 4.3424,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976153772071257e-05,
+      "loss": 4.3441,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967767824560737e-05,
+      "loss": 4.34,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595939825585395e-05,
+      "loss": 4.3403,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595102868714716e-05,
+      "loss": 4.3377,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594264273963664e-05,
+      "loss": 4.3223,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.329,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592588722341933e-05,
+      "loss": 4.322,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591750127590882e-05,
+      "loss": 4.3276,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59091153283983e-05,
+      "loss": 4.328,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590072938088778e-05,
+      "loss": 4.3324,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589234343337726e-05,
+      "loss": 4.3275,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588395748586673e-05,
+      "loss": 4.3197,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587557153835621e-05,
+      "loss": 4.334,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586718559084569e-05,
+      "loss": 4.3297,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.334,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3267,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584204412711786e-05,
+      "loss": 4.3134,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583365817960734e-05,
+      "loss": 4.3306,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582527223209682e-05,
+      "loss": 4.3179,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816902663390035e-05,
+      "loss": 4.3263,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808516715879515e-05,
+      "loss": 4.3165,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.3174,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3164,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3235,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.3179,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3198,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3184,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749831462109604e-05,
+      "loss": 4.3099,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3076,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733059567088564e-05,
+      "loss": 4.3296,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724673619578044e-05,
+      "loss": 4.2979,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716287672067524e-05,
+      "loss": 4.2989,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5707901724557004e-05,
+      "loss": 4.3052,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699515777046484e-05,
+      "loss": 4.322,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691129829535963e-05,
+      "loss": 4.3084,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682743882025443e-05,
+      "loss": 4.3052,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674357934514923e-05,
+      "loss": 4.2991,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56659719870044e-05,
+      "loss": 4.3055,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565758603949388e-05,
+      "loss": 4.3162,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.3089,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564083052327657e-05,
+      "loss": 4.3079,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563244457576605e-05,
+      "loss": 4.3195,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3175,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561568905954874e-05,
+      "loss": 4.2999,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560730311203822e-05,
+      "loss": 4.3036,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559891716452771e-05,
+      "loss": 4.3139,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559053121701719e-05,
+      "loss": 4.2877,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3102,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.2944,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3181,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.3032,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548634237072046e-05,
+      "loss": 4.2929,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540248289561526e-05,
+      "loss": 4.2926,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.3051,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.3115,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.3014,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506704499519446e-05,
+      "loss": 4.2975,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498318552008926e-05,
+      "loss": 4.2925,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489932604498406e-05,
+      "loss": 4.3047,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481546656987886e-05,
+      "loss": 4.2932,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547316070947737e-05,
+      "loss": 4.2941,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546477476196685e-05,
+      "loss": 4.3047,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545638881445633e-05,
+      "loss": 4.3032,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.2975,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543966605584648e-05,
+      "loss": 4.2845,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543128010833596e-05,
+      "loss": 4.2935,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542289416082544e-05,
+      "loss": 4.2899,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541452459211865e-05,
+      "loss": 4.296,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540613864460813e-05,
+      "loss": 4.2934,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539775269709761e-05,
+      "loss": 4.2827,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389366749587095e-05,
+      "loss": 4.2874,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380980802076575e-05,
+      "loss": 4.2927,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372594854566055e-05,
+      "loss": 4.2693,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364208907055535e-05,
+      "loss": 4.2879,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355822959545015e-05,
+      "loss": 4.2738,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347437012034495e-05,
+      "loss": 4.2927,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339051064523975e-05,
+      "loss": 4.2842,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330665117013455e-05,
+      "loss": 4.2869,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532227916950293e-05,
+      "loss": 4.2806,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5313909600796144e-05,
+      "loss": 4.2827,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305523653285624e-05,
+      "loss": 4.2926,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297137705775104e-05,
+      "loss": 4.2763,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528875175826458e-05,
+      "loss": 4.2884,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528038218955779e-05,
+      "loss": 4.2885,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527199624204728e-05,
+      "loss": 4.2852,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526361029453675e-05,
+      "loss": 4.2865,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525522434702623e-05,
+      "loss": 4.2717,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524683839951571e-05,
+      "loss": 4.2903,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523845245200519e-05,
+      "loss": 4.2766,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523006650449467e-05,
+      "loss": 4.286,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.2743,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521331098827736e-05,
+      "loss": 4.2812,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520494141957057e-05,
+      "loss": 4.2793,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519655547206005e-05,
+      "loss": 4.2733,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518816952454953e-05,
+      "loss": 4.2611,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517978357703902e-05,
+      "loss": 4.2725,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51713976295285e-05,
+      "loss": 4.2682,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516302806082171e-05,
+      "loss": 4.2795,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154642113311187e-05,
+      "loss": 4.2889,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146256165800667e-05,
+      "loss": 4.287,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137870218290147e-05,
+      "loss": 4.2582,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129500649583356e-05,
+      "loss": 4.2844,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121114702072836e-05,
+      "loss": 4.2749,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112745133366045e-05,
+      "loss": 4.2773,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104359185855525e-05,
+      "loss": 4.2668,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095973238345005e-05,
+      "loss": 4.2771,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087587290834485e-05,
+      "loss": 4.2678,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507920134332397e-05,
+      "loss": 4.2755,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507081539581345e-05,
+      "loss": 4.2674,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506242944830293e-05,
+      "loss": 4.2726,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505404350079241e-05,
+      "loss": 4.2714,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504565755328189e-05,
+      "loss": 4.2665,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503727160577137e-05,
+      "loss": 4.2696,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502888565826085e-05,
+      "loss": 4.2779,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502049971075033e-05,
+      "loss": 4.262,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501213014204354e-05,
+      "loss": 4.2762,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500374419453302e-05,
+      "loss": 4.2708,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.273791313171387,
+      "eval_runtime": 288.2941,
+      "eval_samples_per_second": 1323.617,
+      "eval_steps_per_second": 41.364,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49953582470225e-05,
+      "loss": 4.2572,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498697229951198e-05,
+      "loss": 4.2588,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497858635200146e-05,
+      "loss": 4.2718,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497020040449093e-05,
+      "loss": 4.2676,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496181445698042e-05,
+      "loss": 4.2808,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49534285094699e-05,
+      "loss": 4.2521,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494504256195938e-05,
+      "loss": 4.2691,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493665661444886e-05,
+      "loss": 4.2514,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492827066693834e-05,
+      "loss": 4.2678,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491988471942782e-05,
+      "loss": 4.2589,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49114987719173e-05,
+      "loss": 4.2664,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490311282440678e-05,
+      "loss": 4.2714,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489474325569999e-05,
+      "loss": 4.254,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488635730818947e-05,
+      "loss": 4.262,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487797136067895e-05,
+      "loss": 4.2534,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486958541316843e-05,
+      "loss": 4.2473,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486121584446164e-05,
+      "loss": 4.258,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485282989695112e-05,
+      "loss": 4.2572,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844443949440604e-05,
+      "loss": 4.2605,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836058001930084e-05,
+      "loss": 4.2674,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827672054419564e-05,
+      "loss": 4.2722,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819286106909044e-05,
+      "loss": 4.2626,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4810900159398524e-05,
+      "loss": 4.2564,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802514211888004e-05,
+      "loss": 4.268,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479412826437748e-05,
+      "loss": 4.2494,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478574231686696e-05,
+      "loss": 4.263,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477735636935644e-05,
+      "loss": 4.2619,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476897042184592e-05,
+      "loss": 4.2507,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47605844743354e-05,
+      "loss": 4.2502,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475223128443234e-05,
+      "loss": 4.2524,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474384533692182e-05,
+      "loss": 4.2522,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47354593894113e-05,
+      "loss": 4.2566,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472707344190078e-05,
+      "loss": 4.2612,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471868749439026e-05,
+      "loss": 4.2564,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471031792568348e-05,
+      "loss": 4.2572,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470194835697669e-05,
+      "loss": 4.254,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469356240946617e-05,
+      "loss": 4.238,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468517646195565e-05,
+      "loss": 4.2489,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467679051444512e-05,
+      "loss": 4.2389,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46684045669346e-05,
+      "loss": 4.2433,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466001861942408e-05,
+      "loss": 4.2448,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465163267191356e-05,
+      "loss": 4.2567,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464324672440304e-05,
+      "loss": 4.2454,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634860776892526e-05,
+      "loss": 4.2379,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626474829382006e-05,
+      "loss": 4.2555,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618088881871486e-05,
+      "loss": 4.2448,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609702934360966e-05,
+      "loss": 4.2513,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601333365654175e-05,
+      "loss": 4.2485,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4592963796947384e-05,
+      "loss": 4.2313,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584577849436864e-05,
+      "loss": 4.2513,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576191901926344e-05,
+      "loss": 4.2361,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567805954415824e-05,
+      "loss": 4.2476,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559420006905304e-05,
+      "loss": 4.2366,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551050438198513e-05,
+      "loss": 4.2362,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454266449068799e-05,
+      "loss": 4.2389,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453427854317748e-05,
+      "loss": 4.2433,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452589259566696e-05,
+      "loss": 4.2378,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451752302696017e-05,
+      "loss": 4.2392,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450915345825338e-05,
+      "loss": 4.2424,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450076751074286e-05,
+      "loss": 4.2313,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449238156323234e-05,
+      "loss": 4.2276,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448399561572182e-05,
+      "loss": 4.2565,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44756096682113e-05,
+      "loss": 4.2169,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446722372070078e-05,
+      "loss": 4.2209,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445883777319026e-05,
+      "loss": 4.2284,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445045182567974e-05,
+      "loss": 4.2483,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444208225697295e-05,
+      "loss": 4.2261,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443369630946243e-05,
+      "loss": 4.2305,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425310361951914e-05,
+      "loss": 4.2213,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416924414441394e-05,
+      "loss": 4.2299,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408538466930874e-05,
+      "loss": 4.2382,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400152519420354e-05,
+      "loss": 4.235,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439178295071356e-05,
+      "loss": 4.2323,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438339700320304e-05,
+      "loss": 4.2405,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437501105569252e-05,
+      "loss": 4.2404,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4366625108182e-05,
+      "loss": 4.2261,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435823916067148e-05,
+      "loss": 4.2258,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4349853213160956e-05,
+      "loss": 4.2415,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341467265650436e-05,
+      "loss": 4.213,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333081318139916e-05,
+      "loss": 4.2326,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432471174943313e-05,
+      "loss": 4.2198,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431634218072635e-05,
+      "loss": 4.2443,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430795623321583e-05,
+      "loss": 4.2293,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429957028570531e-05,
+      "loss": 4.2199,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429118433819479e-05,
+      "loss": 4.2122,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428279839068426e-05,
+      "loss": 4.2337,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274428821977476e-05,
+      "loss": 4.2339,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266042874466956e-05,
+      "loss": 4.2309,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425765692695643e-05,
+      "loss": 4.2234,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424927097944591e-05,
+      "loss": 4.2231,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424088503193539e-05,
+      "loss": 4.2279,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423249908442487e-05,
+      "loss": 4.2198,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422411313691435e-05,
+      "loss": 4.2202,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215727189403836e-05,
+      "loss": 4.2312,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207357620697045e-05,
+      "loss": 4.2286,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198971673186525e-05,
+      "loss": 4.2256,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190585725676005e-05,
+      "loss": 4.2149,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182199778165485e-05,
+      "loss": 4.2223,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4173830209458694e-05,
+      "loss": 4.2149,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41654606407519e-05,
+      "loss": 4.2249,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415707469324138e-05,
+      "loss": 4.2209,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414868874573086e-05,
+      "loss": 4.2167,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414030279822034e-05,
+      "loss": 4.215,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413191685070982e-05,
+      "loss": 4.2167,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41235309031993e-05,
+      "loss": 4.2075,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411514495568879e-05,
+      "loss": 4.2116,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410675900817827e-05,
+      "loss": 4.2039,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409838943947148e-05,
+      "loss": 4.2217,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409000349196096e-05,
+      "loss": 4.2159,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408161754445044e-05,
+      "loss": 4.2138,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407323159693992e-05,
+      "loss": 4.2086,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406486202823313e-05,
+      "loss": 4.2179,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405647608072261e-05,
+      "loss": 4.2191,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404809013321209e-05,
+      "loss": 4.206,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403970418570157e-05,
+      "loss": 4.2187,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403131823819105e-05,
+      "loss": 4.221,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402296504828799e-05,
+      "loss": 4.2162,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401457910077747e-05,
+      "loss": 4.2196,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400619315326695e-05,
+      "loss": 4.2026,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399780720575643e-05,
+      "loss": 4.2183,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398942125824591e-05,
+      "loss": 4.2125,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398103531073539e-05,
+      "loss": 4.2206,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39726657420286e-05,
+      "loss": 4.199,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396427979451808e-05,
+      "loss": 4.2174,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395591022581129e-05,
+      "loss": 4.2134,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394752427830077e-05,
+      "loss": 4.2048,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393913833079025e-05,
+      "loss": 4.1969,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393075238327973e-05,
+      "loss": 4.204,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392236643576921e-05,
+      "loss": 4.1997,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3913996867062426e-05,
+      "loss": 4.2137,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905610919551906e-05,
+      "loss": 4.2183,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897224972041386e-05,
+      "loss": 4.2231,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888839024530866e-05,
+      "loss": 4.1919,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880453077020346e-05,
+      "loss": 4.2161,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2123,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2049,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.2021,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846909286978266e-05,
+      "loss": 4.216,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383852333946774e-05,
+      "loss": 4.2037,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383013739195722e-05,
+      "loss": 4.2069,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38217514444467e-05,
+      "loss": 4.1999,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381336549693618e-05,
+      "loss": 4.2065,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.2071,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796609980718875e-05,
+      "loss": 4.2036,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788224033208355e-05,
+      "loss": 4.2042,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779838085697835e-05,
+      "loss": 4.2115,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.1995,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763082569480524e-05,
+      "loss": 4.2095,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375471300077374e-05,
+      "loss": 4.2093,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.21806001663208,
+      "eval_runtime": 290.8303,
+      "eval_samples_per_second": 1312.074,
+      "eval_steps_per_second": 41.003,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374632705326321e-05,
+      "loss": 4.1957,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373794110575269e-05,
+      "loss": 4.1925,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372955515824217e-05,
+      "loss": 4.2085,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372116921073165e-05,
+      "loss": 4.2044,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371278326322113e-05,
+      "loss": 4.2104,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370439731571061e-05,
+      "loss": 4.1957,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36960113682001e-05,
+      "loss": 4.2029,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368762542068958e-05,
+      "loss": 4.1887,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367923947317906e-05,
+      "loss": 4.2015,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367085352566854e-05,
+      "loss": 4.1983,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366246757815802e-05,
+      "loss": 4.2019,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36540816306475e-05,
+      "loss": 4.2042,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364571206194071e-05,
+      "loss": 4.1953,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363732611443019e-05,
+      "loss": 4.1946,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362894016691967e-05,
+      "loss": 4.1915,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362055421940915e-05,
+      "loss": 4.1818,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361216827189863e-05,
+      "loss": 4.192,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36037823243881e-05,
+      "loss": 4.195,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359539637687758e-05,
+      "loss": 4.1963,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358701042936707e-05,
+      "loss": 4.2063,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3578640860660284e-05,
+      "loss": 4.2088,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357025491314976e-05,
+      "loss": 4.2038,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356186896563924e-05,
+      "loss": 4.1941,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355348301812872e-05,
+      "loss": 4.2058,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35450970706182e-05,
+      "loss": 4.1895,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3536711123107677e-05,
+      "loss": 4.2,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528325175597157e-05,
+      "loss": 4.1994,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3519955606890366e-05,
+      "loss": 4.186,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511569659379846e-05,
+      "loss": 4.1877,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503183711869326e-05,
+      "loss": 4.1946,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3494814143162535e-05,
+      "loss": 4.1908,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348642819565202e-05,
+      "loss": 4.1953,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34780422481415e-05,
+      "loss": 4.2017,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346965630063098e-05,
+      "loss": 4.194,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346127035312046e-05,
+      "loss": 4.1953,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345288440560994e-05,
+      "loss": 4.1949,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344449845809942e-05,
+      "loss": 4.1831,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34361125105889e-05,
+      "loss": 4.1822,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342772656307838e-05,
+      "loss": 4.1797,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341934061556786e-05,
+      "loss": 4.1836,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341095466805734e-05,
+      "loss": 4.1844,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402568720546814e-05,
+      "loss": 4.1981,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339419915184003e-05,
+      "loss": 4.1848,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338581320432951e-05,
+      "loss": 4.18,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337742725681899e-05,
+      "loss": 4.1967,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336904130930847e-05,
+      "loss": 4.1822,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336065536179795e-05,
+      "loss": 4.1961,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352285793091166e-05,
+      "loss": 4.1869,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3343916224384375e-05,
+      "loss": 4.1735,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3335530276873855e-05,
+      "loss": 4.1908,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3327144329363335e-05,
+      "loss": 4.1796,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3318758381852815e-05,
+      "loss": 4.1921,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331037243434229e-05,
+      "loss": 4.1738,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330198648683177e-05,
+      "loss": 4.1798,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329360053932125e-05,
+      "loss": 4.1766,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285230970614464e-05,
+      "loss": 4.1839,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327684502310394e-05,
+      "loss": 4.1826,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268459075593424e-05,
+      "loss": 4.1761,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326008950688664e-05,
+      "loss": 4.1872,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325170355937611e-05,
+      "loss": 4.1767,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324331761186559e-05,
+      "loss": 4.1686,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323493166435507e-05,
+      "loss": 4.2006,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322654571684455e-05,
+      "loss": 4.1599,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321815976933403e-05,
+      "loss": 4.1589,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320977382182351e-05,
+      "loss": 4.1731,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320138787431299e-05,
+      "loss": 4.1933,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319300192680247e-05,
+      "loss": 4.1653,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318461597929195e-05,
+      "loss": 4.1777,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317624641058516e-05,
+      "loss": 4.1628,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316786046307464e-05,
+      "loss": 4.1716,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315947451556412e-05,
+      "loss": 4.1827,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315108856805361e-05,
+      "loss": 4.1783,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314270262054309e-05,
+      "loss": 4.1784,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31343330518363e-05,
+      "loss": 4.1803,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312594710432578e-05,
+      "loss": 4.1858,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311756115681526e-05,
+      "loss": 4.1663,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310917520930474e-05,
+      "loss": 4.1749,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310078926179422e-05,
+      "loss": 4.1813,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092419693087426e-05,
+      "loss": 4.1597,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084033745576906e-05,
+      "loss": 4.1733,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3075647798066386e-05,
+      "loss": 4.1692,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067261850555866e-05,
+      "loss": 4.1851,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3058875903045346e-05,
+      "loss": 4.176,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305050633433856e-05,
+      "loss": 4.165,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304212038682804e-05,
+      "loss": 4.1604,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303373443931752e-05,
+      "loss": 4.1766,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025348491807e-05,
+      "loss": 4.1769,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301697892310021e-05,
+      "loss": 4.1712,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300859297558969e-05,
+      "loss": 4.1744,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300020702807917e-05,
+      "loss": 4.1693,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299182108056865e-05,
+      "loss": 4.1726,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983435133058124e-05,
+      "loss": 4.1652,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975049185547604e-05,
+      "loss": 4.1645,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296667961684082e-05,
+      "loss": 4.18,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29582936693303e-05,
+      "loss": 4.175,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294990772181978e-05,
+      "loss": 4.1695,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2941538153112995e-05,
+      "loss": 4.1608,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2933152205602475e-05,
+      "loss": 4.1711,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292476625809195e-05,
+      "loss": 4.1597,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291638031058143e-05,
+      "loss": 4.1735,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290799436307091e-05,
+      "loss": 4.1674,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289960841556039e-05,
+      "loss": 4.1654,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289122246804987e-05,
+      "loss": 4.1617,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288283652053935e-05,
+      "loss": 4.1558,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287446695183256e-05,
+      "loss": 4.1603,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286608100432204e-05,
+      "loss": 4.1544,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285769505681152e-05,
+      "loss": 4.1561,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849309109301e-05,
+      "loss": 4.1659,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284092316179048e-05,
+      "loss": 4.1662,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2832537214279964e-05,
+      "loss": 4.1618,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824151266769444e-05,
+      "loss": 4.1532,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281578169806265e-05,
+      "loss": 4.1632,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280739575055213e-05,
+      "loss": 4.1681,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279902618184534e-05,
+      "loss": 4.1562,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279064023433482e-05,
+      "loss": 4.1691,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27822542868243e-05,
+      "loss": 4.1671,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277386833931378e-05,
+      "loss": 4.1663,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276548239180326e-05,
+      "loss": 4.1644,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275711282309647e-05,
+      "loss": 4.1543,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274872687558595e-05,
+      "loss": 4.1607,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274034092807543e-05,
+      "loss": 4.1621,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273197135936865e-05,
+      "loss": 4.1686,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2723585411858127e-05,
+      "loss": 4.1503,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715199464347607e-05,
+      "loss": 4.164,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2706813516837087e-05,
+      "loss": 4.1613,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698427569326566e-05,
+      "loss": 4.1557,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690058000619776e-05,
+      "loss": 4.1432,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2681688431912985e-05,
+      "loss": 4.1552,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673302484402465e-05,
+      "loss": 4.1474,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2664916536891945e-05,
+      "loss": 4.1622,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656530589381425e-05,
+      "loss": 4.1662,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2648144641870904e-05,
+      "loss": 4.1686,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2639758694360384e-05,
+      "loss": 4.1421,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263137274684987e-05,
+      "loss": 4.1651,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262298679933935e-05,
+      "loss": 4.1629,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261460085182883e-05,
+      "loss": 4.1548,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260621490431831e-05,
+      "loss": 4.1503,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597828956807784e-05,
+      "loss": 4.1676,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589443009297264e-05,
+      "loss": 4.1493,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258107344059048e-05,
+      "loss": 4.1599,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257268749307996e-05,
+      "loss": 4.148,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256430154556943e-05,
+      "loss": 4.1593,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255591559805891e-05,
+      "loss": 4.1546,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254752965054839e-05,
+      "loss": 4.1536,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253914370303787e-05,
+      "loss": 4.158,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253075775552735e-05,
+      "loss": 4.1558,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252237180801684e-05,
+      "loss": 4.1574,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251400223931005e-05,
+      "loss": 4.1524,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250561629179953e-05,
+      "loss": 4.1624,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.176931858062744,
+      "eval_runtime": 302.6239,
+      "eval_samples_per_second": 1260.941,
+      "eval_steps_per_second": 39.405,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249723034428901e-05,
+      "loss": 4.1497,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248884439677849e-05,
+      "loss": 4.142,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248045844926797e-05,
+      "loss": 4.1562,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247210525936491e-05,
+      "loss": 4.1561,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246371931185439e-05,
+      "loss": 4.1609,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245533336434387e-05,
+      "loss": 4.1464,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244694741683335e-05,
+      "loss": 4.1512,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243856146932283e-05,
+      "loss": 4.1402,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243017552181231e-05,
+      "loss": 4.1565,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2421789574301793e-05,
+      "loss": 4.1489,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241340362679127e-05,
+      "loss": 4.156,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240501767928075e-05,
+      "loss": 4.1539,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239664811057396e-05,
+      "loss": 4.147,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238826216306344e-05,
+      "loss": 4.1519,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237987621555292e-05,
+      "loss": 4.1388,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23714902680424e-05,
+      "loss": 4.1337,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236312069933561e-05,
+      "loss": 4.1447,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235473475182509e-05,
+      "loss": 4.1535,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234634880431457e-05,
+      "loss": 4.1457,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233796285680405e-05,
+      "loss": 4.1635,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232957690929353e-05,
+      "loss": 4.1581,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232119096178301e-05,
+      "loss": 4.1582,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231280501427249e-05,
+      "loss": 4.1472,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230441906676197e-05,
+      "loss": 4.1564,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2296065876858916e-05,
+      "loss": 4.1417,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2287679929348396e-05,
+      "loss": 4.1568,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2279293981837876e-05,
+      "loss": 4.149,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2270908034327356e-05,
+      "loss": 4.1391,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2262522086816836e-05,
+      "loss": 4.1458,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2254136139306316e-05,
+      "loss": 4.147,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224575019179579e-05,
+      "loss": 4.1441,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2237380623089005e-05,
+      "loss": 4.1486,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2228994675578485e-05,
+      "loss": 4.1512,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2220608728067965e-05,
+      "loss": 4.1537,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2212222780557445e-05,
+      "loss": 4.1426,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.220385321185066e-05,
+      "loss": 4.148,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219546726434014e-05,
+      "loss": 4.1434,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218708131682962e-05,
+      "loss": 4.1302,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2178695369319094e-05,
+      "loss": 4.1326,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217032580061231e-05,
+      "loss": 4.1374,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216193985310179e-05,
+      "loss": 4.1397,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215355390559126e-05,
+      "loss": 4.152,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214516795808074e-05,
+      "loss": 4.1404,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213678201057022e-05,
+      "loss": 4.1365,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21283960630597e-05,
+      "loss": 4.1524,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212001011554918e-05,
+      "loss": 4.1371,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211162416803866e-05,
+      "loss": 4.1508,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210325459933188e-05,
+      "loss": 4.1426,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209486865182136e-05,
+      "loss": 4.1226,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208648270431084e-05,
+      "loss": 4.153,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207809675680032e-05,
+      "loss": 4.1312,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20697108092898e-05,
+      "loss": 4.1471,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206132486177928e-05,
+      "loss": 4.1331,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.205295529307249e-05,
+      "loss": 4.1349,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204456934556197e-05,
+      "loss": 4.1279,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.203618339805145e-05,
+      "loss": 4.1365,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.202779745054093e-05,
+      "loss": 4.1428,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201941150303041e-05,
+      "loss": 4.1277,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011041934323616e-05,
+      "loss": 4.1433,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20026559868131e-05,
+      "loss": 4.1365,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199427003930258e-05,
+      "loss": 4.1249,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198588409179206e-05,
+      "loss": 4.1518,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197749814428154e-05,
+      "loss": 4.1205,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196911219677102e-05,
+      "loss": 4.1171,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196074262806423e-05,
+      "loss": 4.1324,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195237305935744e-05,
+      "loss": 4.1436,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194398711184692e-05,
+      "loss": 4.1223,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19356011643364e-05,
+      "loss": 4.1387,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192721521682588e-05,
+      "loss": 4.1166,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191882926931536e-05,
+      "loss": 4.1341,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191044332180484e-05,
+      "loss": 4.136,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.190205737429432e-05,
+      "loss": 4.1335,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18936714267838e-05,
+      "loss": 4.1361,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188528547927328e-05,
+      "loss": 4.1347,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187689953176276e-05,
+      "loss": 4.1432,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186851358425224e-05,
+      "loss": 4.1238,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186012763674172e-05,
+      "loss": 4.1326,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1851774446838665e-05,
+      "loss": 4.1345,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1843388499328145e-05,
+      "loss": 4.1204,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1835002551817625e-05,
+      "loss": 4.1297,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18266166043071e-05,
+      "loss": 4.1279,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181823065679658e-05,
+      "loss": 4.1399,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180984470928606e-05,
+      "loss": 4.1389,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180145876177554e-05,
+      "loss": 4.1171,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1793072814265025e-05,
+      "loss": 4.1202,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1784686866754505e-05,
+      "loss": 4.1347,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776317298047714e-05,
+      "loss": 4.1368,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1767931350537194e-05,
+      "loss": 4.1307,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1759545403026674e-05,
+      "loss": 4.1247,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175117583431988e-05,
+      "loss": 4.1286,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174278988680936e-05,
+      "loss": 4.1271,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173440393929884e-05,
+      "loss": 4.1269,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172603437059205e-05,
+      "loss": 4.1208,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171764842308153e-05,
+      "loss": 4.1351,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170926247557101e-05,
+      "loss": 4.1307,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170087652806049e-05,
+      "loss": 4.1335,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169249058054998e-05,
+      "loss": 4.1203,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168412101184319e-05,
+      "loss": 4.1251,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167573506433267e-05,
+      "loss": 4.121,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166734911682215e-05,
+      "loss": 4.1314,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165896316931163e-05,
+      "loss": 4.1251,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165057722180111e-05,
+      "loss": 4.1202,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164219127429059e-05,
+      "loss": 4.1255,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163380532678007e-05,
+      "loss": 4.1135,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162541937926955e-05,
+      "loss": 4.118,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161704981056276e-05,
+      "loss": 4.115,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160866386305224e-05,
+      "loss": 4.1176,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160027791554172e-05,
+      "loss": 4.1174,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1591891968031197e-05,
+      "loss": 4.1303,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1583506020520677e-05,
+      "loss": 4.1154,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1575120073010156e-05,
+      "loss": 4.1163,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156673412549964e-05,
+      "loss": 4.1179,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1558348177989116e-05,
+      "loss": 4.1257,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1549962230478596e-05,
+      "loss": 4.1158,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154159266177181e-05,
+      "loss": 4.1291,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1533206714261285e-05,
+      "loss": 4.1239,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152485352435823e-05,
+      "loss": 4.1244,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151646757684771e-05,
+      "loss": 4.1249,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150808162933719e-05,
+      "loss": 4.1177,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149969568182667e-05,
+      "loss": 4.1189,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149130973431615e-05,
+      "loss": 4.1236,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.148292378680563e-05,
+      "loss": 4.1217,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147453783929511e-05,
+      "loss": 4.1135,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146615189178459e-05,
+      "loss": 4.1258,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1457782323077806e-05,
+      "loss": 4.122,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1449396375567286e-05,
+      "loss": 4.1173,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144101042805676e-05,
+      "loss": 4.1014,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143262448054624e-05,
+      "loss": 4.1141,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142423853303572e-05,
+      "loss": 4.1068,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14158525855252e-05,
+      "loss": 4.1212,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140746663801468e-05,
+      "loss": 4.1268,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139908069050416e-05,
+      "loss": 4.128,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139071112179737e-05,
+      "loss": 4.1032,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382341553090584e-05,
+      "loss": 4.1251,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1373955605580064e-05,
+      "loss": 4.1205,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1365569658069544e-05,
+      "loss": 4.1195,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357183710559024e-05,
+      "loss": 4.1117,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1348797763048504e-05,
+      "loss": 4.1262,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340411815537984e-05,
+      "loss": 4.1088,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1332025868027464e-05,
+      "loss": 4.1237,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1323639920516944e-05,
+      "loss": 4.1051,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131527035181015e-05,
+      "loss": 4.1197,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130690078310336e-05,
+      "loss": 4.1166,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129851483559284e-05,
+      "loss": 4.1146,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129012888808232e-05,
+      "loss": 4.1174,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12817429405718e-05,
+      "loss": 4.1157,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127337337186502e-05,
+      "loss": 4.1174,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12649874243545e-05,
+      "loss": 4.1146,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125660147684398e-05,
+      "loss": 4.1251,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.146052837371826,
+      "eval_runtime": 300.8643,
+      "eval_samples_per_second": 1268.316,
+      "eval_steps_per_second": 39.636,
+      "step": 534240
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.124821552933346e-05,
+      "loss": 4.11,
+      "step": 534528
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.123982958182294e-05,
+      "loss": 4.1019,
+      "step": 535040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.123144363431242e-05,
+      "loss": 4.1195,
+      "step": 535552
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.12230576868019e-05,
+      "loss": 4.1181,
+      "step": 536064
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.121467173929138e-05,
+      "loss": 4.1183,
+      "step": 536576
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1206302170584586e-05,
+      "loss": 4.1103,
+      "step": 537088
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1197916223074066e-05,
+      "loss": 4.1133,
+      "step": 537600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1189530275563546e-05,
+      "loss": 4.1017,
+      "step": 538112
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1181144328053026e-05,
+      "loss": 4.1142,
+      "step": 538624
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1172758380542506e-05,
+      "loss": 4.113,
+      "step": 539136
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.116438881183572e-05,
+      "loss": 4.1137,
+      "step": 539648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.115601924312893e-05,
+      "loss": 4.116,
+      "step": 540160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.114764967442214e-05,
+      "loss": 4.1123,
+      "step": 540672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.113926372691162e-05,
+      "loss": 4.1082,
+      "step": 541184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.11308777794011e-05,
+      "loss": 4.1015,
+      "step": 541696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.112249183189058e-05,
+      "loss": 4.0944,
+      "step": 542208
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.111410588438006e-05,
+      "loss": 4.1088,
+      "step": 542720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.110571993686954e-05,
+      "loss": 4.1129,
+      "step": 543232
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.109733398935902e-05,
+      "loss": 4.1086,
+      "step": 543744
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.10889480418485e-05,
+      "loss": 4.1194,
+      "step": 544256
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.108056209433798e-05,
+      "loss": 4.1217,
+      "step": 544768
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.107217614682746e-05,
+      "loss": 4.1214,
+      "step": 545280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.106379019931694e-05,
+      "loss": 4.1101,
+      "step": 545792
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1055420630610156e-05,
+      "loss": 4.1122,
+      "step": 546304
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1047034683099636e-05,
+      "loss": 4.1109,
+      "step": 546816
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1038648735589115e-05,
+      "loss": 4.1147,
+      "step": 547328
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1030279166882325e-05,
+      "loss": 4.1103,
+      "step": 547840
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1021893219371805e-05,
+      "loss": 4.1058,
+      "step": 548352
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1013507271861285e-05,
+      "loss": 4.1097,
+      "step": 548864
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.1005121324350764e-05,
+      "loss": 4.1065,
+      "step": 549376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0996735376840244e-05,
+      "loss": 4.1041,
+      "step": 549888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.098834942932972e-05,
+      "loss": 4.1126,
+      "step": 550400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.09799634818192e-05,
+      "loss": 4.1148,
+      "step": 550912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.097157753430868e-05,
+      "loss": 4.1151,
+      "step": 551424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.096320796560189e-05,
+      "loss": 4.1017,
+      "step": 551936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.095482201809137e-05,
+      "loss": 4.1147,
+      "step": 552448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.094643607058085e-05,
+      "loss": 4.1084,
+      "step": 552960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.093805012307033e-05,
+      "loss": 4.0926,
+      "step": 553472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.092968055436354e-05,
+      "loss": 4.0959,
+      "step": 553984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.092129460685302e-05,
+      "loss": 4.0986,
+      "step": 554496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.09129086593425e-05,
+      "loss": 4.1048,
+      "step": 555008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.090452271183198e-05,
+      "loss": 4.1101,
+      "step": 555520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.089613676432146e-05,
+      "loss": 4.1062,
+      "step": 556032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.088775081681094e-05,
+      "loss": 4.0959,
+      "step": 556544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.087936486930042e-05,
+      "loss": 4.119,
+      "step": 557056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.08709789217899e-05,
+      "loss": 4.1003,
+      "step": 557568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.086259297427938e-05,
+      "loss": 4.1149,
+      "step": 558080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.08542234055726e-05,
+      "loss": 4.1073,
+      "step": 558592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.084583745806208e-05,
+      "loss": 4.0814,
+      "step": 559104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.083745151055156e-05,
+      "loss": 4.1214,
+      "step": 559616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.082906556304104e-05,
+      "loss": 4.093,
+      "step": 560128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0820712373137976e-05,
+      "loss": 4.1077,
+      "step": 560640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0812326425627456e-05,
+      "loss": 4.0993,
+      "step": 561152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0803956856920665e-05,
+      "loss": 4.0959,
+      "step": 561664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0795570909410145e-05,
+      "loss": 4.0962,
+      "step": 562176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0787184961899625e-05,
+      "loss": 4.0972,
+      "step": 562688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0778799014389105e-05,
+      "loss": 4.1105,
+      "step": 563200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0770413066878585e-05,
+      "loss": 4.0897,
+      "step": 563712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0762027119368065e-05,
+      "loss": 4.1058,
+      "step": 564224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.075364117185755e-05,
+      "loss": 4.1035,
+      "step": 564736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.074525522434703e-05,
+      "loss": 4.0849,
+      "step": 565248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.073686927683651e-05,
+      "loss": 4.1112,
+      "step": 565760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.072848332932599e-05,
+      "loss": 4.0898,
+      "step": 566272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.07201137606192e-05,
+      "loss": 4.0835,
+      "step": 566784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.071172781310868e-05,
+      "loss": 4.0922,
+      "step": 567296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.070334186559816e-05,
+      "loss": 4.1047,
+      "step": 567808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.069495591808764e-05,
+      "loss": 4.0863,
+      "step": 568320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.068656997057712e-05,
+      "loss": 4.1024,
+      "step": 568832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.06781840230666e-05,
+      "loss": 4.0835,
+      "step": 569344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.066981445435981e-05,
+      "loss": 4.0926,
+      "step": 569856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.066144488565302e-05,
+      "loss": 4.1017,
+      "step": 570368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0653058938142505e-05,
+      "loss": 4.0968,
+      "step": 570880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0644672990631985e-05,
+      "loss": 4.103,
+      "step": 571392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0636287043121465e-05,
+      "loss": 4.0961,
+      "step": 571904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0627901095610945e-05,
+      "loss": 4.106,
+      "step": 572416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0619515148100425e-05,
+      "loss": 4.0924,
+      "step": 572928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0611129200589905e-05,
+      "loss": 4.0987,
+      "step": 573440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.060274325307938e-05,
+      "loss": 4.0926,
+      "step": 573952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0594373684372594e-05,
+      "loss": 4.0893,
+      "step": 574464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0585987736862074e-05,
+      "loss": 4.0948,
+      "step": 574976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0577601789351554e-05,
+      "loss": 4.0938,
+      "step": 575488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.056921584184103e-05,
+      "loss": 4.1029,
+      "step": 576000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.056082989433051e-05,
+      "loss": 4.1075,
+      "step": 576512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.055244394681999e-05,
+      "loss": 4.0787,
+      "step": 577024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.05440743781132e-05,
+      "loss": 4.0839,
+      "step": 577536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.053568843060268e-05,
+      "loss": 4.0962,
+      "step": 578048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.05273188618959e-05,
+      "loss": 4.103,
+      "step": 578560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.051893291438538e-05,
+      "loss": 4.0954,
+      "step": 579072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.051054696687485e-05,
+      "loss": 4.0901,
+      "step": 579584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.050216101936433e-05,
+      "loss": 4.0986,
+      "step": 580096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.049377507185381e-05,
+      "loss": 4.0904,
+      "step": 580608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.048538912434329e-05,
+      "loss": 4.0929,
+      "step": 581120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.047700317683277e-05,
+      "loss": 4.0865,
+      "step": 581632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.046861722932225e-05,
+      "loss": 4.0978,
+      "step": 582144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.046023128181173e-05,
+      "loss": 4.0991,
+      "step": 582656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.045186171310494e-05,
+      "loss": 4.0988,
+      "step": 583168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.044347576559443e-05,
+      "loss": 4.0885,
+      "step": 583680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.043508981808391e-05,
+      "loss": 4.0882,
+      "step": 584192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.042670387057339e-05,
+      "loss": 4.0856,
+      "step": 584704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.041831792306287e-05,
+      "loss": 4.0987,
+      "step": 585216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.040993197555235e-05,
+      "loss": 4.0948,
+      "step": 585728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.040154602804183e-05,
+      "loss": 4.085,
+      "step": 586240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.039316008053131e-05,
+      "loss": 4.0909,
+      "step": 586752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0384790511824516e-05,
+      "loss": 4.0822,
+      "step": 587264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0376404564313996e-05,
+      "loss": 4.0824,
+      "step": 587776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0368034995607205e-05,
+      "loss": 4.0787,
+      "step": 588288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0359649048096685e-05,
+      "loss": 4.0844,
+      "step": 588800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0351263100586165e-05,
+      "loss": 4.0856,
+      "step": 589312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0342877153075645e-05,
+      "loss": 4.0926,
+      "step": 589824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0334491205565125e-05,
+      "loss": 4.0858,
+      "step": 590336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.032612163685834e-05,
+      "loss": 4.0752,
+      "step": 590848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.031773568934782e-05,
+      "loss": 4.0871,
+      "step": 591360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.03093497418373e-05,
+      "loss": 4.0905,
+      "step": 591872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.030096379432678e-05,
+      "loss": 4.0873,
+      "step": 592384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.029257784681626e-05,
+      "loss": 4.0916,
+      "step": 592896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.028420827810947e-05,
+      "loss": 4.0879,
+      "step": 593408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.027583870940268e-05,
+      "loss": 4.0973,
+      "step": 593920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.026745276189216e-05,
+      "loss": 4.0872,
+      "step": 594432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.025906681438164e-05,
+      "loss": 4.084,
+      "step": 594944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.025068086687112e-05,
+      "loss": 4.0827,
+      "step": 595456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.02422949193606e-05,
+      "loss": 4.0979,
+      "step": 595968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.023390897185008e-05,
+      "loss": 4.0864,
+      "step": 596480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.022552302433956e-05,
+      "loss": 4.0793,
+      "step": 596992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.021713707682904e-05,
+      "loss": 4.0904,
+      "step": 597504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.020875112931852e-05,
+      "loss": 4.0931,
+      "step": 598016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0200365181808e-05,
+      "loss": 4.0814,
+      "step": 598528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.019197923429748e-05,
+      "loss": 4.072,
+      "step": 599040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.018360966559069e-05,
+      "loss": 4.08,
+      "step": 599552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.017522371808017e-05,
+      "loss": 4.0742,
+      "step": 600064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.016683777056965e-05,
+      "loss": 4.0875,
+      "step": 600576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.015845182305913e-05,
+      "loss": 4.0934,
+      "step": 601088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.015006587554861e-05,
+      "loss": 4.0955,
+      "step": 601600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0141696306841817e-05,
+      "loss": 4.069,
+      "step": 602112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0133310359331297e-05,
+      "loss": 4.0941,
+      "step": 602624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.012492441182078e-05,
+      "loss": 4.0862,
+      "step": 603136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.011653846431026e-05,
+      "loss": 4.0874,
+      "step": 603648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.010815251679974e-05,
+      "loss": 4.0789,
+      "step": 604160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.009976656928922e-05,
+      "loss": 4.0898,
+      "step": 604672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.00913806217787e-05,
+      "loss": 4.0779,
+      "step": 605184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.008299467426818e-05,
+      "loss": 4.0935,
+      "step": 605696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.007462510556139e-05,
+      "loss": 4.0734,
+      "step": 606208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.006623915805087e-05,
+      "loss": 4.0847,
+      "step": 606720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.005785321054035e-05,
+      "loss": 4.0847,
+      "step": 607232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.004946726302983e-05,
+      "loss": 4.0838,
+      "step": 607744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.004109769432304e-05,
+      "loss": 4.0825,
+      "step": 608256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.003271174681252e-05,
+      "loss": 4.082,
+      "step": 608768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.0024325799302e-05,
+      "loss": 4.0867,
+      "step": 609280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.001593985179148e-05,
+      "loss": 4.0826,
+      "step": 609792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.00075702830847e-05,
+      "loss": 4.091,
+      "step": 610304
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.121993064880371,
+      "eval_runtime": 301.019,
+      "eval_samples_per_second": 1267.664,
+      "eval_steps_per_second": 39.615,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999918433557418e-05,
+      "loss": 4.0758,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999079838806366e-05,
+      "loss": 4.0739,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998241244055314e-05,
+      "loss": 4.0836,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997402649304262e-05,
+      "loss": 4.0845,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99656405455321e-05,
+      "loss": 4.0911,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957254598021577e-05,
+      "loss": 4.0782,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994886865051105e-05,
+      "loss": 4.0852,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994048270300053e-05,
+      "loss": 4.0714,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993209675549001e-05,
+      "loss": 4.0785,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992371080797949e-05,
+      "loss": 4.0812,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991532486046897e-05,
+      "loss": 4.083,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990693891295845e-05,
+      "loss": 4.0841,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9898569344251665e-05,
+      "loss": 4.0846,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890183396741145e-05,
+      "loss": 4.0733,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9881797449230625e-05,
+      "loss": 4.0712,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873411501720105e-05,
+      "loss": 4.0628,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865025554209585e-05,
+      "loss": 4.0764,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9856639606699065e-05,
+      "loss": 4.0853,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848253659188545e-05,
+      "loss": 4.0744,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9839867711678025e-05,
+      "loss": 4.0873,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9831481764167505e-05,
+      "loss": 4.0901,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823112195460714e-05,
+      "loss": 4.0885,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9814726247950194e-05,
+      "loss": 4.0839,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9806340300439674e-05,
+      "loss": 4.0759,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9797954352929154e-05,
+      "loss": 4.0785,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9789568405418634e-05,
+      "loss": 4.0816,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781182457908114e-05,
+      "loss": 4.0859,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9772796510397594e-05,
+      "loss": 4.0757,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9764410562887074e-05,
+      "loss": 4.0773,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975604099418029e-05,
+      "loss": 4.076,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974765504666976e-05,
+      "loss": 4.0706,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973926909915924e-05,
+      "loss": 4.0811,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973088315164872e-05,
+      "loss": 4.0823,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97224972041382e-05,
+      "loss": 4.0845,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971412763543141e-05,
+      "loss": 4.073,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970575806672463e-05,
+      "loss": 4.0818,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969737211921411e-05,
+      "loss": 4.0791,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968898617170359e-05,
+      "loss": 4.0603,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968060022419307e-05,
+      "loss": 4.0689,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672230655486283e-05,
+      "loss": 4.0667,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9663844707975763e-05,
+      "loss": 4.0697,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965547513926897e-05,
+      "loss": 4.0854,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964708919175845e-05,
+      "loss": 4.0749,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963870324424793e-05,
+      "loss": 4.0691,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9630317296737406e-05,
+      "loss": 4.0814,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9621931349226886e-05,
+      "loss": 4.0704,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613545401716366e-05,
+      "loss": 4.0855,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605159454205845e-05,
+      "loss": 4.0769,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959678988549906e-05,
+      "loss": 4.0509,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958840393798854e-05,
+      "loss": 4.0897,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958001799047802e-05,
+      "loss": 4.0635,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95716320429675e-05,
+      "loss": 4.0768,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956324609545698e-05,
+      "loss": 4.0677,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955486014794646e-05,
+      "loss": 4.0676,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954647420043594e-05,
+      "loss": 4.0657,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953810463172915e-05,
+      "loss": 4.0682,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952971868421863e-05,
+      "loss": 4.0835,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952133273670811e-05,
+      "loss": 4.057,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951294678919759e-05,
+      "loss": 4.0811,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950456084168707e-05,
+      "loss": 4.0708,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949617489417655e-05,
+      "loss": 4.0528,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948778894666603e-05,
+      "loss": 4.0839,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947940299915551e-05,
+      "loss": 4.0635,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471033430448726e-05,
+      "loss": 4.0512,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462647482938206e-05,
+      "loss": 4.0652,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454261535427686e-05,
+      "loss": 4.075,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445875587917166e-05,
+      "loss": 4.0605,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437489640406646e-05,
+      "loss": 4.0707,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429103692896126e-05,
+      "loss": 4.057,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420734124189335e-05,
+      "loss": 4.0608,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9412348176678815e-05,
+      "loss": 4.0717,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9403962229168295e-05,
+      "loss": 4.0697,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939557628165777e-05,
+      "loss": 4.0705,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938719033414725e-05,
+      "loss": 4.0682,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937880438663673e-05,
+      "loss": 4.0789,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370418439126214e-05,
+      "loss": 4.062,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362032491615694e-05,
+      "loss": 4.0697,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935367930171264e-05,
+      "loss": 4.062,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934529335420212e-05,
+      "loss": 4.0601,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933692378549533e-05,
+      "loss": 4.0668,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932853783798481e-05,
+      "loss": 4.0614,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932015189047429e-05,
+      "loss": 4.0727,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931176594296377e-05,
+      "loss": 4.0816,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930337999545324e-05,
+      "loss": 4.0498,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929499404794272e-05,
+      "loss": 4.0551,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92866081004322e-05,
+      "loss": 4.0685,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927822215292168e-05,
+      "loss": 4.0698,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926983620541116e-05,
+      "loss": 4.0678,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926145025790065e-05,
+      "loss": 4.0615,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925306431039013e-05,
+      "loss": 4.0688,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924469474168334e-05,
+      "loss": 4.0623,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923630879417282e-05,
+      "loss": 4.0616,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92279228466623e-05,
+      "loss": 4.0608,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921953689915178e-05,
+      "loss": 4.0715,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211167330444986e-05,
+      "loss": 4.0723,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9202781382934466e-05,
+      "loss": 4.0695,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9194395435423946e-05,
+      "loss": 4.0556,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9186009487913426e-05,
+      "loss": 4.061,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9177623540402906e-05,
+      "loss": 4.0584,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169237592892386e-05,
+      "loss": 4.0673,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9160851645381866e-05,
+      "loss": 4.0682,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9152465697871346e-05,
+      "loss": 4.0538,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914407975036083e-05,
+      "loss": 4.0646,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913571018165404e-05,
+      "loss": 4.0513,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912732423414352e-05,
+      "loss": 4.0553,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9118938286633e-05,
+      "loss": 4.0505,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911055233912248e-05,
+      "loss": 4.0609,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910218277041569e-05,
+      "loss": 4.0592,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909379682290517e-05,
+      "loss": 4.0633,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908541087539465e-05,
+      "loss": 4.056,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907702492788413e-05,
+      "loss": 4.0527,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906865535917734e-05,
+      "loss": 4.0556,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906026941166682e-05,
+      "loss": 4.066,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90518834641563e-05,
+      "loss": 4.0575,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9043497516645786e-05,
+      "loss": 4.0611,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903511156913526e-05,
+      "loss": 4.0608,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026742000428475e-05,
+      "loss": 4.0706,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018356052917955e-05,
+      "loss": 4.0603,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900997010540743e-05,
+      "loss": 4.0565,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900158415789691e-05,
+      "loss": 4.0601,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899319821038639e-05,
+      "loss": 4.0642,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898481226287587e-05,
+      "loss": 4.0585,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897642631536535e-05,
+      "loss": 4.0531,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896804036785483e-05,
+      "loss": 4.0605,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895965442034431e-05,
+      "loss": 4.0655,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895126847283379e-05,
+      "loss": 4.0542,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894288252532327e-05,
+      "loss": 4.0446,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8934496577812755e-05,
+      "loss": 4.0521,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926110630302235e-05,
+      "loss": 4.0459,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891775744039917e-05,
+      "loss": 4.0639,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890937149288865e-05,
+      "loss": 4.0621,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890098554537813e-05,
+      "loss": 4.0679,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889259959786761e-05,
+      "loss": 4.0426,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888423002916082e-05,
+      "loss": 4.0664,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88758440816503e-05,
+      "loss": 4.06,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886745813413978e-05,
+      "loss": 4.0607,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885907218662926e-05,
+      "loss": 4.0548,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885068623911874e-05,
+      "loss": 4.0552,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884230029160822e-05,
+      "loss": 4.0525,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88339143440977e-05,
+      "loss": 4.0676,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882552839658719e-05,
+      "loss": 4.0445,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881714244907667e-05,
+      "loss": 4.0587,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880877288036988e-05,
+      "loss": 4.0596,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8800403311663086e-05,
+      "loss": 4.06,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792017364152566e-05,
+      "loss": 4.0499,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8783631416642046e-05,
+      "loss": 4.0562,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8775245469131526e-05,
+      "loss": 4.0598,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8766875900424735e-05,
+      "loss": 4.0554,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8758489952914215e-05,
+      "loss": 4.0637,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.102855205535889,
+      "eval_runtime": 298.3708,
+      "eval_samples_per_second": 1278.915,
+      "eval_steps_per_second": 39.967,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750104005403695e-05,
+      "loss": 4.0495,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8741718057893175e-05,
+      "loss": 4.0438,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8733332110382655e-05,
+      "loss": 4.0559,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872496254167587e-05,
+      "loss": 4.0591,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871657659416535e-05,
+      "loss": 4.0622,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870819064665483e-05,
+      "loss": 4.0526,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869980469914431e-05,
+      "loss": 4.0601,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869141875163379e-05,
+      "loss": 4.0426,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683032804123264e-05,
+      "loss": 4.0522,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8674646856612744e-05,
+      "loss": 4.0544,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8666260909102224e-05,
+      "loss": 4.0542,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8657874961591704e-05,
+      "loss": 4.0569,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864950539288491e-05,
+      "loss": 4.0623,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864111944537439e-05,
+      "loss": 4.0451,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863273349786388e-05,
+      "loss": 4.0444,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862434755035336e-05,
+      "loss": 4.0343,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861596160284284e-05,
+      "loss": 4.0543,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860759203413605e-05,
+      "loss": 4.0591,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859920608662553e-05,
+      "loss": 4.0468,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859082013911501e-05,
+      "loss": 4.0585,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858243419160449e-05,
+      "loss": 4.0681,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857404824409397e-05,
+      "loss": 4.0616,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856566229658345e-05,
+      "loss": 4.0565,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855727634907293e-05,
+      "loss": 4.0517,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854890678036614e-05,
+      "loss": 4.0562,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854052083285562e-05,
+      "loss": 4.0517,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85321348853451e-05,
+      "loss": 4.0577,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852374893783458e-05,
+      "loss": 4.0553,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8515362990324064e-05,
+      "loss": 4.0483,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8506977042813544e-05,
+      "loss": 4.0457,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8498591095303024e-05,
+      "loss": 4.0455,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8490205147792504e-05,
+      "loss": 4.0559,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848183557908571e-05,
+      "loss": 4.0583,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847344963157519e-05,
+      "loss": 4.0577,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846506368406467e-05,
+      "loss": 4.0498,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845669411535788e-05,
+      "loss": 4.0504,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844830816784736e-05,
+      "loss": 4.0545,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843992222033684e-05,
+      "loss": 4.0355,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843155265163005e-05,
+      "loss": 4.0418,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842316670411953e-05,
+      "loss": 4.0424,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841478075660902e-05,
+      "loss": 4.0421,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84063948090985e-05,
+      "loss": 4.0582,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839800886158798e-05,
+      "loss": 4.0474,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838962291407745e-05,
+      "loss": 4.0433,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838123696656693e-05,
+      "loss": 4.0562,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837285101905641e-05,
+      "loss": 4.0422,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836446507154589e-05,
+      "loss": 4.055,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835607912403537e-05,
+      "loss": 4.0603,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834769317652485e-05,
+      "loss": 4.026,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833932360781806e-05,
+      "loss": 4.0578,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833093766030754e-05,
+      "loss": 4.0423,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832255171279702e-05,
+      "loss": 4.0513,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83141657652865e-05,
+      "loss": 4.0436,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8305796196579716e-05,
+      "loss": 4.0375,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8297410249069195e-05,
+      "loss": 4.0428,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8289040680362405e-05,
+      "loss": 4.0402,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8280654732851885e-05,
+      "loss": 4.0594,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8272268785341364e-05,
+      "loss": 4.0317,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8263882837830844e-05,
+      "loss": 4.0571,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8255496890320324e-05,
+      "loss": 4.0464,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8247110942809804e-05,
+      "loss": 4.028,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8238724995299284e-05,
+      "loss": 4.0582,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230339047788764e-05,
+      "loss": 4.0347,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822196947908197e-05,
+      "loss": 4.0268,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821358353157145e-05,
+      "loss": 4.0361,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820519758406093e-05,
+      "loss": 4.0531,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819681163655042e-05,
+      "loss": 4.0324,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81884256890399e-05,
+      "loss": 4.0487,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818003974152938e-05,
+      "loss": 4.0325,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817165379401886e-05,
+      "loss": 4.0327,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816326784650834e-05,
+      "loss": 4.0433,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815489827780155e-05,
+      "loss": 4.0453,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814651233029103e-05,
+      "loss": 4.0468,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813814276158424e-05,
+      "loss": 4.045,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812975681407372e-05,
+      "loss": 4.0502,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81213708665632e-05,
+      "loss": 4.0422,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811298491905268e-05,
+      "loss": 4.0438,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810459897154216e-05,
+      "loss": 4.0305,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809621302403164e-05,
+      "loss": 4.0389,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808782707652112e-05,
+      "loss": 4.0426,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80794411290106e-05,
+      "loss": 4.0335,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071071560303814e-05,
+      "loss": 4.048,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806268561279329e-05,
+      "loss": 4.0536,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805429966528277e-05,
+      "loss": 4.032,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804591371777225e-05,
+      "loss": 4.0284,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037527770261727e-05,
+      "loss": 4.0426,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029141822751207e-05,
+      "loss": 4.0449,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8020755875240687e-05,
+      "loss": 4.0429,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801240268533763e-05,
+      "loss": 4.0347,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800401673782711e-05,
+      "loss": 4.0458,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799564716912033e-05,
+      "loss": 4.0377,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798726122160981e-05,
+      "loss": 4.0384,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797887527409929e-05,
+      "loss": 4.0367,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797048932658876e-05,
+      "loss": 4.0465,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796210337907824e-05,
+      "loss": 4.0442,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795371743156772e-05,
+      "loss": 4.0459,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79453314840572e-05,
+      "loss": 4.0283,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793694553654668e-05,
+      "loss": 4.0392,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792855958903616e-05,
+      "loss": 4.0315,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792017364152564e-05,
+      "loss": 4.0434,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791178769401512e-05,
+      "loss": 4.0423,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79034017465046e-05,
+      "loss": 4.033,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789503217779781e-05,
+      "loss": 4.0401,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7886646230287296e-05,
+      "loss": 4.0261,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7878276661580505e-05,
+      "loss": 4.033,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7869890714069985e-05,
+      "loss": 4.0257,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7861504766559465e-05,
+      "loss": 4.0344,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7853118819048945e-05,
+      "loss": 4.035,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7844732871538425e-05,
+      "loss": 4.0374,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836346924027905e-05,
+      "loss": 4.0315,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7827960976517385e-05,
+      "loss": 4.03,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7819575029006865e-05,
+      "loss": 4.0321,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7811205460300074e-05,
+      "loss": 4.0391,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7802819512789554e-05,
+      "loss": 4.0351,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7794433565279034e-05,
+      "loss": 4.0353,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7786047617768514e-05,
+      "loss": 4.0354,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777767804906173e-05,
+      "loss": 4.0439,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776930848035494e-05,
+      "loss": 4.0384,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776092253284442e-05,
+      "loss": 4.0328,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775255296413763e-05,
+      "loss": 4.0342,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774416701662711e-05,
+      "loss": 4.0406,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773578106911659e-05,
+      "loss": 4.0374,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772739512160607e-05,
+      "loss": 4.0303,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771900917409555e-05,
+      "loss": 4.0301,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771062322658503e-05,
+      "loss": 4.0433,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770223727907451e-05,
+      "loss": 4.0297,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769385133156399e-05,
+      "loss": 4.0209,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76854817628572e-05,
+      "loss": 4.0274,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767711219415041e-05,
+      "loss": 4.0201,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766872624663989e-05,
+      "loss": 4.0449,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766034029912937e-05,
+      "loss": 4.0324,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765195435161885e-05,
+      "loss": 4.0481,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764356840410833e-05,
+      "loss": 4.023,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763518245659781e-05,
+      "loss": 4.0409,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762679650908729e-05,
+      "loss": 4.0361,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761841056157677e-05,
+      "loss": 4.0382,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761004099286998e-05,
+      "loss": 4.0319,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760165504535946e-05,
+      "loss": 4.0309,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759326909784894e-05,
+      "loss": 4.0262,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758489952914215e-05,
+      "loss": 4.0466,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757651358163164e-05,
+      "loss": 4.0217,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756812763412112e-05,
+      "loss": 4.0367,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75597416866106e-05,
+      "loss": 4.0397,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755135573910007e-05,
+      "loss": 4.0333,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754296979158955e-05,
+      "loss": 4.0276,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753458384407903e-05,
+      "loss": 4.0318,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752619789656851e-05,
+      "loss": 4.0372,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751782832786172e-05,
+      "loss": 4.035,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75094423803512e-05,
+      "loss": 4.04,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.086784839630127,
+      "eval_runtime": 299.0054,
+      "eval_samples_per_second": 1276.201,
+      "eval_steps_per_second": 39.882,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750105643284068e-05,
+      "loss": 4.0205,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749267048533016e-05,
+      "loss": 4.0211,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748428453781964e-05,
+      "loss": 4.0332,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747589859030912e-05,
+      "loss": 4.0357,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467512642798605e-05,
+      "loss": 4.0415,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459126695288085e-05,
+      "loss": 4.0303,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7450740747777565e-05,
+      "loss": 4.0341,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7442354800267045e-05,
+      "loss": 4.0238,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7433968852756525e-05,
+      "loss": 4.027,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7425582905246005e-05,
+      "loss": 4.0292,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417196957735485e-05,
+      "loss": 4.0292,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740881101022496e-05,
+      "loss": 4.0376,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7400441441518174e-05,
+      "loss": 4.0383,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7392055494007654e-05,
+      "loss": 4.0229,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738366954649713e-05,
+      "loss": 4.0228,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737528359898661e-05,
+      "loss": 4.0107,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736691403027982e-05,
+      "loss": 4.0344,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73585280827693e-05,
+      "loss": 4.0311,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735014213525878e-05,
+      "loss": 4.0244,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734175618774826e-05,
+      "loss": 4.033,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733338661904148e-05,
+      "loss": 4.0476,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732500067153096e-05,
+      "loss": 4.0392,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731661472402043e-05,
+      "loss": 4.0336,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730822877650991e-05,
+      "loss": 4.0285,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729984282899939e-05,
+      "loss": 4.0371,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729145688148887e-05,
+      "loss": 4.0283,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728307093397835e-05,
+      "loss": 4.0293,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727468498646783e-05,
+      "loss": 4.033,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726631541776104e-05,
+      "loss": 4.0277,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725792947025053e-05,
+      "loss": 4.0268,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724954352274001e-05,
+      "loss": 4.0244,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724115757522949e-05,
+      "loss": 4.0323,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723277162771897e-05,
+      "loss": 4.0328,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722440205901218e-05,
+      "loss": 4.0357,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7216016111501657e-05,
+      "loss": 4.0282,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7207630163991137e-05,
+      "loss": 4.0273,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7199244216480616e-05,
+      "loss": 4.0338,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7190858268970096e-05,
+      "loss": 4.0176,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7182488700263306e-05,
+      "loss": 4.0198,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7174102752752785e-05,
+      "loss": 4.0196,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7165716805242265e-05,
+      "loss": 4.0197,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157330857731745e-05,
+      "loss": 4.0327,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7148944910221225e-05,
+      "loss": 4.0271,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140558962710705e-05,
+      "loss": 4.0293,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713217301520019e-05,
+      "loss": 4.0276,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712378706768967e-05,
+      "loss": 4.0233,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711541749898288e-05,
+      "loss": 4.0344,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710703155147236e-05,
+      "loss": 4.0351,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709866198276557e-05,
+      "loss": 4.0046,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709027603525505e-05,
+      "loss": 4.0358,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708189008774453e-05,
+      "loss": 4.0191,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707350414023401e-05,
+      "loss": 4.0319,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706511819272349e-05,
+      "loss": 4.0251,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705673224521296e-05,
+      "loss": 4.0148,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704834629770244e-05,
+      "loss": 4.0237,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703996035019193e-05,
+      "loss": 4.0143,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703157440268141e-05,
+      "loss": 4.0339,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702320483397462e-05,
+      "loss": 4.0132,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70148188864641e-05,
+      "loss": 4.0324,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700643293895358e-05,
+      "loss": 4.0278,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699804699144306e-05,
+      "loss": 4.0115,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698966104393254e-05,
+      "loss": 4.0332,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698129147522575e-05,
+      "loss": 4.0122,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697290552771523e-05,
+      "loss": 4.0082,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696451958020471e-05,
+      "loss": 4.0121,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695613363269419e-05,
+      "loss": 4.0309,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69477640639874e-05,
+      "loss": 4.0119,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6939378116476884e-05,
+      "loss": 4.0307,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6930992168966363e-05,
+      "loss": 4.0056,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922606221455843e-05,
+      "loss": 4.0165,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691423665274905e-05,
+      "loss": 4.0233,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690585070523853e-05,
+      "loss": 4.0214,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689748113653174e-05,
+      "loss": 4.0267,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688909518902122e-05,
+      "loss": 4.0253,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68807092415107e-05,
+      "loss": 4.0285,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687233967280391e-05,
+      "loss": 4.0178,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686395372529339e-05,
+      "loss": 4.0232,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685556777778287e-05,
+      "loss": 4.0133,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684718183027235e-05,
+      "loss": 4.0186,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683879588276184e-05,
+      "loss": 4.0183,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683040993525132e-05,
+      "loss": 4.0119,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68220239877408e-05,
+      "loss": 4.0262,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681363804023028e-05,
+      "loss": 4.0337,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6805268471523486e-05,
+      "loss": 4.0167,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6796882524012966e-05,
+      "loss": 4.0024,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6788496576502446e-05,
+      "loss": 4.0232,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6780110628991926e-05,
+      "loss": 4.023,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6771724681481406e-05,
+      "loss": 4.0276,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6763338733970886e-05,
+      "loss": 4.0112,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6754952786460366e-05,
+      "loss": 4.0257,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6746566838949846e-05,
+      "loss": 4.0152,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738197270243055e-05,
+      "loss": 4.0154,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729811322732535e-05,
+      "loss": 4.0137,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672142537522202e-05,
+      "loss": 4.0238,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67130394277115e-05,
+      "loss": 4.0229,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670466985900471e-05,
+      "loss": 4.026,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669628391149419e-05,
+      "loss": 4.0086,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668789796398367e-05,
+      "loss": 4.0163,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667951201647315e-05,
+      "loss": 4.0121,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6671126068962624e-05,
+      "loss": 4.0221,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662740121452104e-05,
+      "loss": 4.021,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665437055274532e-05,
+      "loss": 4.0154,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66459846052348e-05,
+      "loss": 4.017,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663759865772427e-05,
+      "loss": 4.0049,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662921271021376e-05,
+      "loss": 4.0165,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662082676270324e-05,
+      "loss": 4.0015,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661244081519272e-05,
+      "loss": 4.0135,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660407124648593e-05,
+      "loss": 4.0119,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659568529897541e-05,
+      "loss": 4.0201,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658729935146489e-05,
+      "loss": 4.0091,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657891340395437e-05,
+      "loss": 4.0101,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657052745644385e-05,
+      "loss": 4.0086,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656215788773706e-05,
+      "loss": 4.0216,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655377194022654e-05,
+      "loss": 4.0154,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654538599271602e-05,
+      "loss": 4.0145,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65370000452055e-05,
+      "loss": 4.018,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652861409769498e-05,
+      "loss": 4.0241,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652022815018446e-05,
+      "loss": 4.0152,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651185858147767e-05,
+      "loss": 4.0132,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650348901277088e-05,
+      "loss": 4.0137,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649510306526036e-05,
+      "loss": 4.02,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648671711774984e-05,
+      "loss": 4.0153,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647833117023932e-05,
+      "loss": 4.013,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64699452227288e-05,
+      "loss": 4.013,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646155927521828e-05,
+      "loss": 4.0232,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645317332770776e-05,
+      "loss": 4.0063,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644478738019724e-05,
+      "loss": 4.0061,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643641781149045e-05,
+      "loss": 4.0013,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6428064621587396e-05,
+      "loss": 4.0057,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6419678674076876e-05,
+      "loss": 4.0208,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6411292726566356e-05,
+      "loss": 4.0172,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6402906779055836e-05,
+      "loss": 4.0287,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6394520831545316e-05,
+      "loss": 4.0005,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6386134884034796e-05,
+      "loss": 4.0201,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6377748936524276e-05,
+      "loss": 4.013,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6369362989013756e-05,
+      "loss": 4.0177,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6360977041503236e-05,
+      "loss": 4.0108,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6352591093992715e-05,
+      "loss": 4.0098,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6344205146482195e-05,
+      "loss": 4.0121,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6335819198971675e-05,
+      "loss": 4.0269,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6327433251461155e-05,
+      "loss": 3.9972,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6319047303950635e-05,
+      "loss": 4.0187,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310661356440115e-05,
+      "loss": 4.0172,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630229178773333e-05,
+      "loss": 4.0116,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629390584022281e-05,
+      "loss": 4.0097,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6285519892712284e-05,
+      "loss": 4.0135,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6277133945201764e-05,
+      "loss": 4.0141,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6268747997691244e-05,
+      "loss": 4.018,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626037842898446e-05,
+      "loss": 4.0174,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.074056625366211,
+      "eval_runtime": 288.995,
+      "eval_samples_per_second": 1320.407,
+      "eval_steps_per_second": 41.264,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625200886027767e-05,
+      "loss": 4.0045,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624362291276715e-05,
+      "loss": 4.0034,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623523696525663e-05,
+      "loss": 4.0136,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622685101774611e-05,
+      "loss": 4.0148,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621846507023558e-05,
+      "loss": 4.0229,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621007912272507e-05,
+      "loss": 4.0113,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6201709554018285e-05,
+      "loss": 4.0163,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619332360650776e-05,
+      "loss": 4.0075,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618493765899724e-05,
+      "loss": 4.0032,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617655171148672e-05,
+      "loss": 4.0067,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61681657639762e-05,
+      "loss": 4.014,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615977981646568e-05,
+      "loss": 4.0138,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615141024775889e-05,
+      "loss": 4.0219,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614302430024837e-05,
+      "loss": 4.0025,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613463835273785e-05,
+      "loss": 4.0036,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612625240522733e-05,
+      "loss": 3.9905,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611786645771681e-05,
+      "loss": 4.0141,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610948051020629e-05,
+      "loss": 4.0126,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610109456269577e-05,
+      "loss": 4.0016,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609270861518525e-05,
+      "loss": 4.0141,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608432266767473e-05,
+      "loss": 4.0319,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607593672016421e-05,
+      "loss": 4.0172,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606756715145742e-05,
+      "loss": 4.0149,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60591812039469e-05,
+      "loss": 4.0102,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605079525643638e-05,
+      "loss": 4.0172,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604242568772959e-05,
+      "loss": 4.0114,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603403974021907e-05,
+      "loss": 4.0109,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602565379270855e-05,
+      "loss": 4.0096,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601726784519803e-05,
+      "loss": 4.0101,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600888189768751e-05,
+      "loss": 4.0062,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600049595017699e-05,
+      "loss": 4.0088,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599211000266647e-05,
+      "loss": 4.0078,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598372405515595e-05,
+      "loss": 4.0127,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597533810764543e-05,
+      "loss": 4.0192,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596696853893865e-05,
+      "loss": 4.0096,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595858259142812e-05,
+      "loss": 4.007,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59501966439176e-05,
+      "loss": 4.0167,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594181069640708e-05,
+      "loss": 3.9975,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593342474889656e-05,
+      "loss": 4.002,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592505518018977e-05,
+      "loss": 3.9977,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591666923267925e-05,
+      "loss": 4.0025,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5908299663972465e-05,
+      "loss": 4.0078,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5899913716461945e-05,
+      "loss": 4.0075,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5891527768951425e-05,
+      "loss": 4.012,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5883141821440905e-05,
+      "loss": 4.0074,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5874755873930385e-05,
+      "loss": 4.0083,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5866369926419865e-05,
+      "loss": 4.0116,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5858000357713074e-05,
+      "loss": 4.0172,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5849614410202554e-05,
+      "loss": 3.9874,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5841228462692034e-05,
+      "loss": 4.0171,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5832842515181514e-05,
+      "loss": 3.9994,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582447294647472e-05,
+      "loss": 4.0111,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58160869989642e-05,
+      "loss": 4.0056,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580770105145368e-05,
+      "loss": 3.9981,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579931510394316e-05,
+      "loss": 3.9986,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579092915643264e-05,
+      "loss": 3.9968,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578254320892212e-05,
+      "loss": 4.0199,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577415726141161e-05,
+      "loss": 3.992,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576577131390109e-05,
+      "loss": 4.0143,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57574017451943e-05,
+      "loss": 4.0025,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574901579768378e-05,
+      "loss": 3.9946,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574062985017326e-05,
+      "loss": 4.0111,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573224390266274e-05,
+      "loss": 3.9981,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572385795515222e-05,
+      "loss": 3.9869,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57154720076417e-05,
+      "loss": 3.9941,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570708606013118e-05,
+      "loss": 4.0153,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569871649142439e-05,
+      "loss": 3.9947,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569033054391387e-05,
+      "loss": 4.012,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568194459640335e-05,
+      "loss": 3.9863,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567357502769656e-05,
+      "loss": 3.997,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566518908018604e-05,
+      "loss": 4.0043,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565680313267552e-05,
+      "loss": 4.0069,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5648417185165e-05,
+      "loss": 4.0036,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564003123765448e-05,
+      "loss": 4.0071,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5631645290143956e-05,
+      "loss": 4.0104,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5623259342633436e-05,
+      "loss": 3.9989,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5614873395122916e-05,
+      "loss": 4.0022,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5606487447612396e-05,
+      "loss": 3.996,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5598101500101876e-05,
+      "loss": 4.0053,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5589715552591356e-05,
+      "loss": 3.9979,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5581329605080836e-05,
+      "loss": 3.9959,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5572960036374045e-05,
+      "loss": 4.0081,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556457408886353e-05,
+      "loss": 4.0129,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555618814135301e-05,
+      "loss": 3.9999,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554781857264622e-05,
+      "loss": 3.9815,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55394326251357e-05,
+      "loss": 4.0038,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553104667762518e-05,
+      "loss": 4.0046,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552266073011466e-05,
+      "loss": 4.0089,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551427478260414e-05,
+      "loss": 3.9938,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550588883509362e-05,
+      "loss": 4.0027,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54975028875831e-05,
+      "loss": 4.0063,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548911694007258e-05,
+      "loss": 3.9941,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548074737136579e-05,
+      "loss": 3.9929,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.547236142385527e-05,
+      "loss": 4.009,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546397547634475e-05,
+      "loss": 4.0013,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5455605907637965e-05,
+      "loss": 4.0099,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5447219960127445e-05,
+      "loss": 3.9948,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5438834012616925e-05,
+      "loss": 3.9959,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5430448065106405e-05,
+      "loss": 3.9927,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5422062117595885e-05,
+      "loss": 4.0062,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5413676170085365e-05,
+      "loss": 4.0047,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5405290222574845e-05,
+      "loss": 3.9939,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539690427506432e-05,
+      "loss": 4.0009,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5388534706357534e-05,
+      "loss": 3.9891,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538016513765074e-05,
+      "loss": 3.9974,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537177919014022e-05,
+      "loss": 3.9825,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53633932426297e-05,
+      "loss": 3.9951,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535500729511918e-05,
+      "loss": 3.9937,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534662134760866e-05,
+      "loss": 4.0013,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533823540009814e-05,
+      "loss": 3.995,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532984945258762e-05,
+      "loss": 3.9985,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53214635050771e-05,
+      "loss": 3.9869,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531309393637032e-05,
+      "loss": 4.0034,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530472436766353e-05,
+      "loss": 3.9969,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529633842015301e-05,
+      "loss": 3.9966,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528795247264249e-05,
+      "loss": 3.9973,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527956652513197e-05,
+      "loss": 4.0093,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5271196956425177e-05,
+      "loss": 3.9954,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5262811008914657e-05,
+      "loss": 3.9949,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5254425061404136e-05,
+      "loss": 3.9992,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5246039113893616e-05,
+      "loss": 4.0021,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5237653166383096e-05,
+      "loss": 3.9959,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522928359767631e-05,
+      "loss": 3.9997,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522089765016579e-05,
+      "loss": 3.9931,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5212511702655265e-05,
+      "loss": 4.001,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520414213394848e-05,
+      "loss": 3.9863,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519575618643796e-05,
+      "loss": 3.9901,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5187370238927434e-05,
+      "loss": 3.9839,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517900067022065e-05,
+      "loss": 3.9914,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517061472271013e-05,
+      "loss": 4.0003,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516224515400334e-05,
+      "loss": 3.9978,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5153859206492826e-05,
+      "loss": 4.0082,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145473258982306e-05,
+      "loss": 3.9875,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137087311471786e-05,
+      "loss": 3.9983,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5128701363961266e-05,
+      "loss": 4.0016,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512031541645074e-05,
+      "loss": 3.9991,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511192946894022e-05,
+      "loss": 3.9935,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51035435214297e-05,
+      "loss": 3.989,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509515757391918e-05,
+      "loss": 3.9965,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508678800521239e-05,
+      "loss": 4.0025,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507840205770187e-05,
+      "loss": 3.9873,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507001611019135e-05,
+      "loss": 3.9986,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506163016268083e-05,
+      "loss": 4.0002,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505324421517031e-05,
+      "loss": 3.9958,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5044858267659795e-05,
+      "loss": 3.9925,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5036472320149275e-05,
+      "loss": 3.9957,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5028086372638755e-05,
+      "loss": 3.9947,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5019700425128234e-05,
+      "loss": 3.9974,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5011330856421444e-05,
+      "loss": 4.0037,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.063039302825928,
+      "eval_runtime": 292.9068,
+      "eval_samples_per_second": 1302.773,
+      "eval_steps_per_second": 40.713,
+      "step": 915840
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.6671459159038086e+17,
+  "trial_name": null,
+  "trial_params": null
+}