diff --git "a/checkpoint-2213280/trainer_state.json" "b/checkpoint-2213280/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2213280/trainer_state.json"
@@ -0,0 +1,26189 @@
+{
+  "best_metric": 3.997385263442993,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/lstm/1/checkpoints/checkpoint-2213280",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 2213280,
+  "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
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5002944908910924e-05,
+      "loss": 3.9896,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4994558961400404e-05,
+      "loss": 3.9838,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4986173013889883e-05,
+      "loss": 3.9949,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4977787066379363e-05,
+      "loss": 3.9995,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496940111886884e-05,
+      "loss": 4.0022,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496101517135832e-05,
+      "loss": 3.9961,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4952629223847797e-05,
+      "loss": 3.9955,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4944243276337276e-05,
+      "loss": 3.9921,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493585732882676e-05,
+      "loss": 3.9868,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492747138131624e-05,
+      "loss": 3.9908,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491908543380572e-05,
+      "loss": 3.9989,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49106994862952e-05,
+      "loss": 3.9954,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490232991758841e-05,
+      "loss": 4.005,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489394397007789e-05,
+      "loss": 3.9824,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488555802256737e-05,
+      "loss": 3.9871,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487717207505685e-05,
+      "loss": 3.9749,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486878612754633e-05,
+      "loss": 3.998,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486040018003581e-05,
+      "loss": 3.9946,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485201423252529e-05,
+      "loss": 3.9903,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484362828501477e-05,
+      "loss": 3.9906,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483524233750425e-05,
+      "loss": 4.0124,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482685638999373e-05,
+      "loss": 4.0016,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481847044248321e-05,
+      "loss": 3.9969,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481008449497269e-05,
+      "loss": 3.9906,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480171492626591e-05,
+      "loss": 4.0044,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479332897875539e-05,
+      "loss": 3.9927,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4784959410048597e-05,
+      "loss": 3.9894,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4776573462538077e-05,
+      "loss": 4.0001,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4768203893831286e-05,
+      "loss": 3.9861,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4759817946320766e-05,
+      "loss": 3.9927,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4751431998810246e-05,
+      "loss": 3.9923,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4743046051299726e-05,
+      "loss": 3.9893,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4734660103789205e-05,
+      "loss": 3.996,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726274156278685e-05,
+      "loss": 4.0048,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717888208768165e-05,
+      "loss": 3.9913,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4709502261257645e-05,
+      "loss": 3.9904,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470113269255086e-05,
+      "loss": 4.0001,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469276312384407e-05,
+      "loss": 3.9797,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468437717633355e-05,
+      "loss": 3.9848,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467599122882303e-05,
+      "loss": 3.9867,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466760528131251e-05,
+      "loss": 3.9831,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465921933380199e-05,
+      "loss": 3.9901,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465083338629146e-05,
+      "loss": 3.9917,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464244743878094e-05,
+      "loss": 3.9936,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463406149127042e-05,
+      "loss": 3.9908,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46256755437599e-05,
+      "loss": 3.9901,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461728959624938e-05,
+      "loss": 3.9976,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460890364873886e-05,
+      "loss": 3.9979,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460051770122835e-05,
+      "loss": 3.9766,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459216451132529e-05,
+      "loss": 3.997,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458377856381477e-05,
+      "loss": 3.9825,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457539261630425e-05,
+      "loss": 3.9934,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456700666879373e-05,
+      "loss": 3.9901,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455862072128321e-05,
+      "loss": 3.9806,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455023477377269e-05,
+      "loss": 3.9824,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45418652050659e-05,
+      "loss": 3.9779,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453347925755538e-05,
+      "loss": 4.0067,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452509331004486e-05,
+      "loss": 3.9768,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451670736253434e-05,
+      "loss": 3.9954,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450832141502382e-05,
+      "loss": 3.9874,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449995184631703e-05,
+      "loss": 3.9789,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449156589880651e-05,
+      "loss": 3.9898,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448317995129599e-05,
+      "loss": 3.9841,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447479400378547e-05,
+      "loss": 3.9732,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446642443507868e-05,
+      "loss": 3.9731,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445803848756816e-05,
+      "loss": 4.0023,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444965254005764e-05,
+      "loss": 3.9796,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444129935015458e-05,
+      "loss": 3.9873,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443291340264406e-05,
+      "loss": 3.9715,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442452745513354e-05,
+      "loss": 3.9841,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4416141507623026e-05,
+      "loss": 3.9821,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4407755560112506e-05,
+      "loss": 3.9923,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4399369612601986e-05,
+      "loss": 3.9892,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4390983665091466e-05,
+      "loss": 3.99,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4382597717580946e-05,
+      "loss": 3.993,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4374211770070426e-05,
+      "loss": 3.9805,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4365825822559906e-05,
+      "loss": 3.9848,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4357439875049386e-05,
+      "loss": 3.9805,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4349053927538866e-05,
+      "loss": 3.9882,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340667980028346e-05,
+      "loss": 3.9845,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433228203251782e-05,
+      "loss": 3.9761,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43238960850073e-05,
+      "loss": 3.9939,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431551013749678e-05,
+      "loss": 3.996,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430712418998626e-05,
+      "loss": 3.985,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4298754621279475e-05,
+      "loss": 3.963,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429038505257269e-05,
+      "loss": 3.9862,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428199910506217e-05,
+      "loss": 3.9903,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427361315755165e-05,
+      "loss": 3.9945,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4265227210041124e-05,
+      "loss": 3.9822,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425685764133434e-05,
+      "loss": 3.9854,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424847169382382e-05,
+      "loss": 3.9886,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424008574631329e-05,
+      "loss": 3.9774,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423169979880277e-05,
+      "loss": 3.9788,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422331385129225e-05,
+      "loss": 3.9898,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421492790378173e-05,
+      "loss": 3.982,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420654195627121e-05,
+      "loss": 3.9966,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419815600876069e-05,
+      "loss": 3.9804,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418977006125017e-05,
+      "loss": 3.9791,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418140049254339e-05,
+      "loss": 3.9776,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417301454503287e-05,
+      "loss": 3.9886,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416462859752235e-05,
+      "loss": 3.9906,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415624265001183e-05,
+      "loss": 3.9781,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414785670250131e-05,
+      "loss": 3.9829,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413948713379452e-05,
+      "loss": 3.9729,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131117565087726e-05,
+      "loss": 3.9847,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122731617577206e-05,
+      "loss": 3.9621,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114345670066686e-05,
+      "loss": 3.9786,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4105959722556166e-05,
+      "loss": 3.982,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409759015384938e-05,
+      "loss": 3.986,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408920420633886e-05,
+      "loss": 3.9736,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408081825882834e-05,
+      "loss": 3.9814,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407243231131782e-05,
+      "loss": 3.9727,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406406274261103e-05,
+      "loss": 3.9867,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405569317390424e-05,
+      "loss": 3.9796,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404730722639372e-05,
+      "loss": 3.9789,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40389212788832e-05,
+      "loss": 3.9794,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403053533137268e-05,
+      "loss": 3.9935,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402214938386216e-05,
+      "loss": 3.9808,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401376343635164e-05,
+      "loss": 3.9788,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400537748884112e-05,
+      "loss": 3.9836,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39969915413306e-05,
+      "loss": 3.9814,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398860559382008e-05,
+      "loss": 3.9804,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398021964630957e-05,
+      "loss": 3.9814,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397183369879905e-05,
+      "loss": 3.9765,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3963447751288527e-05,
+      "loss": 3.986,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3955078182581736e-05,
+      "loss": 3.9691,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3946692235071216e-05,
+      "loss": 3.9788,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3938322666364425e-05,
+      "loss": 3.9688,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929953097657634e-05,
+      "loss": 3.979,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921567150147114e-05,
+      "loss": 3.9771,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913181202636594e-05,
+      "loss": 3.9855,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904795255126074e-05,
+      "loss": 3.9923,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896409307615554e-05,
+      "loss": 3.9731,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3888023360105034e-05,
+      "loss": 3.9817,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387963741259452e-05,
+      "loss": 3.9819,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871251465084e-05,
+      "loss": 3.9825,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386288189637721e-05,
+      "loss": 3.9816,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385449594886669e-05,
+      "loss": 3.9728,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384611000135617e-05,
+      "loss": 3.9839,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383772405384565e-05,
+      "loss": 3.9882,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382933810633513e-05,
+      "loss": 3.971,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38209521588246e-05,
+      "loss": 3.9822,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381256621131408e-05,
+      "loss": 3.9862,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38041966426073e-05,
+      "loss": 3.9752,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379581069509678e-05,
+      "loss": 3.9775,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378742474758626e-05,
+      "loss": 3.983,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377903880007574e-05,
+      "loss": 3.9772,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377065285256522e-05,
+      "loss": 3.9796,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37622669050547e-05,
+      "loss": 3.9901,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.053833961486816,
+      "eval_runtime": 288.8733,
+      "eval_samples_per_second": 1320.963,
+      "eval_steps_per_second": 41.281,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375388095754418e-05,
+      "loss": 3.9793,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374549501003366e-05,
+      "loss": 3.964,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373710906252314e-05,
+      "loss": 3.979,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372872311501262e-05,
+      "loss": 3.9858,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.37203371675021e-05,
+      "loss": 3.9826,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371195121999158e-05,
+      "loss": 3.9831,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370356527248106e-05,
+      "loss": 3.9772,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369517932497054e-05,
+      "loss": 3.9774,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368680975626375e-05,
+      "loss": 3.9711,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367842380875323e-05,
+      "loss": 3.9789,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367003786124271e-05,
+      "loss": 3.9797,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3661651913732187e-05,
+      "loss": 3.9799,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36532823450254e-05,
+      "loss": 3.9919,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.364489639751488e-05,
+      "loss": 3.9672,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.363651045000436e-05,
+      "loss": 3.9697,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362812450249384e-05,
+      "loss": 3.9592,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3619738554983316e-05,
+      "loss": 3.9828,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3611352607472795e-05,
+      "loss": 3.979,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3602966659962275e-05,
+      "loss": 3.9757,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3594580712451755e-05,
+      "loss": 3.9753,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3586194764941235e-05,
+      "loss": 3.9972,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3577808817430715e-05,
+      "loss": 3.9839,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3569422869920195e-05,
+      "loss": 3.9809,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3561036922409675e-05,
+      "loss": 3.9749,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3552650974899155e-05,
+      "loss": 3.9927,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3544265027388635e-05,
+      "loss": 3.9732,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353587907987812e-05,
+      "loss": 3.9775,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35274931323676e-05,
+      "loss": 3.9858,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351910718485708e-05,
+      "loss": 3.9714,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351072123734656e-05,
+      "loss": 3.9747,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3502335289836035e-05,
+      "loss": 3.9774,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3493949342325515e-05,
+      "loss": 3.9703,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348557977361873e-05,
+      "loss": 3.9839,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347721020491194e-05,
+      "loss": 3.9877,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346882425740142e-05,
+      "loss": 3.9752,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346045468869463e-05,
+      "loss": 3.975,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345206874118411e-05,
+      "loss": 3.9881,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344368279367359e-05,
+      "loss": 3.9582,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3435296846163076e-05,
+      "loss": 3.9733,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3426910898652555e-05,
+      "loss": 3.9726,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341852495114203e-05,
+      "loss": 3.9637,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341013900363151e-05,
+      "loss": 3.9763,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340175305612099e-05,
+      "loss": 3.9808,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.339336710861047e-05,
+      "loss": 3.9784,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338498116109995e-05,
+      "loss": 3.9747,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337659521358943e-05,
+      "loss": 3.9729,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336822564488264e-05,
+      "loss": 3.9761,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3359856076175853e-05,
+      "loss": 3.985,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3351470128665327e-05,
+      "loss": 3.9633,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334308418115481e-05,
+      "loss": 3.9813,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333469823364429e-05,
+      "loss": 3.9726,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332631228613377e-05,
+      "loss": 3.9749,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331792633862325e-05,
+      "loss": 3.9757,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330954039111273e-05,
+      "loss": 3.962,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330115444360221e-05,
+      "loss": 3.9734,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329276849609169e-05,
+      "loss": 3.9583,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.328438254858117e-05,
+      "loss": 3.9916,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.327599660107065e-05,
+      "loss": 3.9613,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326761065356013e-05,
+      "loss": 3.9831,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325922470604961e-05,
+      "loss": 3.9681,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325087151614655e-05,
+      "loss": 3.9641,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324248556863603e-05,
+      "loss": 3.9707,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323409962112551e-05,
+      "loss": 3.9715,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322571367361499e-05,
+      "loss": 3.962,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321734410490821e-05,
+      "loss": 3.9543,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320895815739769e-05,
+      "loss": 3.9827,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320057220988717e-05,
+      "loss": 3.9699,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.319218626237665e-05,
+      "loss": 3.9719,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3183816693669856e-05,
+      "loss": 3.9572,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3175430746159336e-05,
+      "loss": 3.9637,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3167044798648816e-05,
+      "loss": 3.9701,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3158658851138296e-05,
+      "loss": 3.976,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3150272903627776e-05,
+      "loss": 3.9738,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3141886956117256e-05,
+      "loss": 3.9763,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3133501008606736e-05,
+      "loss": 3.9779,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3125115061096215e-05,
+      "loss": 3.9659,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3116729113585695e-05,
+      "loss": 3.9691,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310837592368264e-05,
+      "loss": 3.9679,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309998997617212e-05,
+      "loss": 3.9717,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30916040286616e-05,
+      "loss": 3.9698,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308321808115108e-05,
+      "loss": 3.9618,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307483213364056e-05,
+      "loss": 3.9773,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306644618613004e-05,
+      "loss": 3.9812,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3058060238619513e-05,
+      "loss": 3.9667,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304967429110899e-05,
+      "loss": 3.952,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304130472240221e-05,
+      "loss": 3.9723,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303291877489169e-05,
+      "loss": 3.9736,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302453282738117e-05,
+      "loss": 3.9761,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301614687987065e-05,
+      "loss": 3.9655,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300776093236013e-05,
+      "loss": 3.9692,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299939136365334e-05,
+      "loss": 3.9735,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299100541614282e-05,
+      "loss": 3.9617,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29826194686323e-05,
+      "loss": 3.9658,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297423352112178e-05,
+      "loss": 3.9712,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296584757361126e-05,
+      "loss": 3.9682,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295747800490447e-05,
+      "loss": 3.9812,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294909205739395e-05,
+      "loss": 3.9698,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294070610988343e-05,
+      "loss": 3.9599,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293232016237291e-05,
+      "loss": 3.9621,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292393421486239e-05,
+      "loss": 3.973,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291554826735187e-05,
+      "loss": 3.9745,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2907162319841354e-05,
+      "loss": 3.9629,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289879275113456e-05,
+      "loss": 3.9651,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289040680362404e-05,
+      "loss": 3.9588,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.288202085611352e-05,
+      "loss": 3.9723,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2873634908603e-05,
+      "loss": 3.9459,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286526533989621e-05,
+      "loss": 3.9613,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285689577118942e-05,
+      "loss": 3.9659,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28485098236789e-05,
+      "loss": 3.9718,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284012387616838e-05,
+      "loss": 3.9603,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283175430746159e-05,
+      "loss": 3.9662,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2823368359951076e-05,
+      "loss": 3.9579,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2814982412440556e-05,
+      "loss": 3.969,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2806596464930036e-05,
+      "loss": 3.9692,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2798210517419516e-05,
+      "loss": 3.9612,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2789824569908996e-05,
+      "loss": 3.9666,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2781438622398476e-05,
+      "loss": 3.9807,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2773069053691685e-05,
+      "loss": 3.9647,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2764683106181165e-05,
+      "loss": 3.9662,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2756297158670645e-05,
+      "loss": 3.9657,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2747927589963854e-05,
+      "loss": 3.966,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2739541642453334e-05,
+      "loss": 3.9675,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2731155694942814e-05,
+      "loss": 3.9678,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2722769747432294e-05,
+      "loss": 3.9638,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2714383799921774e-05,
+      "loss": 3.9704,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270599785241126e-05,
+      "loss": 3.9583,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269761190490074e-05,
+      "loss": 3.96,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268922595739022e-05,
+      "loss": 3.9573,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26808400098797e-05,
+      "loss": 3.9622,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267247044117291e-05,
+      "loss": 3.9556,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266408449366239e-05,
+      "loss": 3.9764,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265569854615187e-05,
+      "loss": 3.9769,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264731259864135e-05,
+      "loss": 3.9568,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263892665113082e-05,
+      "loss": 3.9683,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26305407036203e-05,
+      "loss": 3.9695,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262217113491352e-05,
+      "loss": 3.9696,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2613785187403e-05,
+      "loss": 3.9587,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260539923989248e-05,
+      "loss": 3.9598,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259701329238196e-05,
+      "loss": 3.971,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2588643723675175e-05,
+      "loss": 3.9727,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258025777616465e-05,
+      "loss": 3.956,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257187182865413e-05,
+      "loss": 3.9674,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.256348588114361e-05,
+      "loss": 3.9743,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255509993363309e-05,
+      "loss": 3.9618,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254671398612257e-05,
+      "loss": 3.9614,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253832803861205e-05,
+      "loss": 3.9686,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252994209110153e-05,
+      "loss": 3.9609,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252155614359101e-05,
+      "loss": 3.9702,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.251317019608049e-05,
+      "loss": 3.9695,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.045386791229248,
+      "eval_runtime": 330.1772,
+      "eval_samples_per_second": 1155.716,
+      "eval_steps_per_second": 36.117,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.250478424856997e-05,
+      "loss": 3.9628,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249639830105945e-05,
+      "loss": 3.9516,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248802873235266e-05,
+      "loss": 3.9613,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247964278484214e-05,
+      "loss": 3.977,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247125683733162e-05,
+      "loss": 3.9682,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24628708898211e-05,
+      "loss": 3.9667,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.245450132111431e-05,
+      "loss": 3.9636,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.244611537360379e-05,
+      "loss": 3.9647,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243772942609327e-05,
+      "loss": 3.956,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242934347858275e-05,
+      "loss": 3.9654,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242097390987596e-05,
+      "loss": 3.9617,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.241260434116917e-05,
+      "loss": 3.9722,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2404234772462386e-05,
+      "loss": 3.9738,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2395848824951866e-05,
+      "loss": 3.9537,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2387462877441346e-05,
+      "loss": 3.956,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2379076929930826e-05,
+      "loss": 3.9436,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2370690982420306e-05,
+      "loss": 3.9669,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2362305034909786e-05,
+      "loss": 3.9659,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2353919087399266e-05,
+      "loss": 3.9616,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2345549518692475e-05,
+      "loss": 3.9584,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2337163571181955e-05,
+      "loss": 3.985,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2328777623671435e-05,
+      "loss": 3.9724,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2320391676160915e-05,
+      "loss": 3.9658,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2312005728650395e-05,
+      "loss": 3.9621,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2303619781139875e-05,
+      "loss": 3.9786,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2295233833629355e-05,
+      "loss": 3.9597,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2286847886118835e-05,
+      "loss": 3.9598,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2278461938608314e-05,
+      "loss": 3.9717,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2270075991097794e-05,
+      "loss": 3.9555,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2261690043587274e-05,
+      "loss": 3.9594,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2253320474880483e-05,
+      "loss": 3.9615,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2244934527369963e-05,
+      "loss": 3.9583,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2236548579859443e-05,
+      "loss": 3.9699,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222816263234892e-05,
+      "loss": 3.9739,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.22197766848384e-05,
+      "loss": 3.9587,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221140711613161e-05,
+      "loss": 3.9695,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.220303754742483e-05,
+      "loss": 3.9683,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.219465159991431e-05,
+      "loss": 3.9464,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.218626565240379e-05,
+      "loss": 3.9582,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217787970489327e-05,
+      "loss": 3.9601,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216949375738275e-05,
+      "loss": 3.9466,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216110780987223e-05,
+      "loss": 3.967,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.215272186236171e-05,
+      "loss": 3.9659,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.214433591485119e-05,
+      "loss": 3.9618,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.213594996734067e-05,
+      "loss": 3.961,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.212756401983015e-05,
+      "loss": 3.9632,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211917807231963e-05,
+      "loss": 3.9587,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211080850361284e-05,
+      "loss": 3.9732,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210242255610232e-05,
+      "loss": 3.9478,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20940366085918e-05,
+      "loss": 3.9645,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.208565066108128e-05,
+      "loss": 3.9658,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207726471357076e-05,
+      "loss": 3.9607,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206887876606024e-05,
+      "loss": 3.9613,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206049281854972e-05,
+      "loss": 3.9464,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20521068710392e-05,
+      "loss": 3.9629,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2043720923528677e-05,
+      "loss": 3.944,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203535135482189e-05,
+      "loss": 3.9772,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202696540731137e-05,
+      "loss": 3.9478,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201859583860458e-05,
+      "loss": 3.9662,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201020989109406e-05,
+      "loss": 3.9582,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.200182394358354e-05,
+      "loss": 3.9525,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1993437996073015e-05,
+      "loss": 3.9561,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1985052048562495e-05,
+      "loss": 3.9607,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.197669885865944e-05,
+      "loss": 3.944,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1968312911148926e-05,
+      "loss": 3.9427,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1959926963638406e-05,
+      "loss": 3.9702,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1951541016127886e-05,
+      "loss": 3.9577,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1943155068617366e-05,
+      "loss": 3.9551,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1934769121106846e-05,
+      "loss": 3.9425,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.192638317359632e-05,
+      "loss": 3.9548,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19179972260858e-05,
+      "loss": 3.9564,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190961127857528e-05,
+      "loss": 3.9637,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190122533106476e-05,
+      "loss": 3.9625,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189285576235797e-05,
+      "loss": 3.9634,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188446981484745e-05,
+      "loss": 3.9618,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187608386733693e-05,
+      "loss": 3.9542,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186769791982641e-05,
+      "loss": 3.9543,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1859311972315895e-05,
+      "loss": 3.9538,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1850926024805375e-05,
+      "loss": 3.9603,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1842540077294855e-05,
+      "loss": 3.9569,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1834154129784335e-05,
+      "loss": 3.9501,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182580093988127e-05,
+      "loss": 3.9606,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.181741499237075e-05,
+      "loss": 3.9665,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.180902904486023e-05,
+      "loss": 3.9565,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.180064309734971e-05,
+      "loss": 3.9449,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.179227352864292e-05,
+      "loss": 3.9525,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.17838875811324e-05,
+      "loss": 3.9596,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.177550163362188e-05,
+      "loss": 3.9654,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176711568611136e-05,
+      "loss": 3.9541,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.175872973860085e-05,
+      "loss": 3.9558,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.175034379109033e-05,
+      "loss": 3.9587,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174195784357981e-05,
+      "loss": 3.9523,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173357189606929e-05,
+      "loss": 3.9497,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.17252023273625e-05,
+      "loss": 3.9576,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.171681637985198e-05,
+      "loss": 3.9608,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170843043234146e-05,
+      "loss": 3.9634,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170004448483094e-05,
+      "loss": 3.962,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169165853732042e-05,
+      "loss": 3.947,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1683305347417356e-05,
+      "loss": 3.9523,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1674919399906836e-05,
+      "loss": 3.9567,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1666533452396315e-05,
+      "loss": 3.9638,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16581475048858e-05,
+      "loss": 3.9522,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164976155737528e-05,
+      "loss": 3.9514,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164137560986476e-05,
+      "loss": 3.9454,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.163300604115797e-05,
+      "loss": 3.9576,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.162462009364745e-05,
+      "loss": 3.9354,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161623414613693e-05,
+      "loss": 3.9503,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160784819862641e-05,
+      "loss": 3.9548,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159946225111589e-05,
+      "loss": 3.9556,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15910926824091e-05,
+      "loss": 3.9486,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.158270673489858e-05,
+      "loss": 3.9524,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157432078738806e-05,
+      "loss": 3.9456,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156593483987754e-05,
+      "loss": 3.9548,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155754889236702e-05,
+      "loss": 3.9573,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15491629448565e-05,
+      "loss": 3.9485,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154077699734598e-05,
+      "loss": 3.9562,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153239104983546e-05,
+      "loss": 3.9641,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152400510232494e-05,
+      "loss": 3.9575,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.151563553361815e-05,
+      "loss": 3.9465,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150724958610763e-05,
+      "loss": 3.9572,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149886363859711e-05,
+      "loss": 3.9542,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149047769108659e-05,
+      "loss": 3.9493,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.148209174357607e-05,
+      "loss": 3.9572,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.147370579606555e-05,
+      "loss": 3.9491,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.146531984855503e-05,
+      "loss": 3.9583,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.145695027984824e-05,
+      "loss": 3.9468,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144856433233772e-05,
+      "loss": 3.9485,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1440178384827204e-05,
+      "loss": 3.9426,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1431808816120413e-05,
+      "loss": 3.9485,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1423422868609893e-05,
+      "loss": 3.9423,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1415036921099373e-05,
+      "loss": 3.9638,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140665097358885e-05,
+      "loss": 3.9653,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139826502607833e-05,
+      "loss": 3.9489,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138987907856781e-05,
+      "loss": 3.9514,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138150950986102e-05,
+      "loss": 3.9588,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13731235623505e-05,
+      "loss": 3.9569,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.136473761483998e-05,
+      "loss": 3.9488,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135635166732946e-05,
+      "loss": 3.9466,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134796571981894e-05,
+      "loss": 3.9589,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133957977230842e-05,
+      "loss": 3.9582,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13311938247979e-05,
+      "loss": 3.9461,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.132280787728739e-05,
+      "loss": 3.9493,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.131442192977687e-05,
+      "loss": 3.9622,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.130603598226634e-05,
+      "loss": 3.9459,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129766641355956e-05,
+      "loss": 3.9502,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128928046604904e-05,
+      "loss": 3.9583,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128089451853851e-05,
+      "loss": 3.9488,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127250857102799e-05,
+      "loss": 3.9573,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126412262351747e-05,
+      "loss": 3.9562,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.038480758666992,
+      "eval_runtime": 300.6748,
+      "eval_samples_per_second": 1269.115,
+      "eval_steps_per_second": 39.661,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125573667600695e-05,
+      "loss": 3.9266,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124735072849643e-05,
+      "loss": 3.9421,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123896478098591e-05,
+      "loss": 3.9475,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123057883347539e-05,
+      "loss": 3.9621,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122219288596487e-05,
+      "loss": 3.9547,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121380693845436e-05,
+      "loss": 3.9612,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120542099094384e-05,
+      "loss": 3.9457,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119703504343332e-05,
+      "loss": 3.9561,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11886490959228e-05,
+      "loss": 3.9403,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118026314841228e-05,
+      "loss": 3.9534,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117187720090176e-05,
+      "loss": 3.9483,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116349125339123e-05,
+      "loss": 3.9617,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.115510530588071e-05,
+      "loss": 3.9629,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114671935837019e-05,
+      "loss": 3.941,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113833341085967e-05,
+      "loss": 3.9449,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112994746334915e-05,
+      "loss": 3.9346,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112156151583863e-05,
+      "loss": 3.9469,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111317556832811e-05,
+      "loss": 3.9529,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110478962081759e-05,
+      "loss": 3.9526,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109640367330707e-05,
+      "loss": 3.9446,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1088034104600286e-05,
+      "loss": 3.9754,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1079648157089766e-05,
+      "loss": 3.9573,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1071262209579246e-05,
+      "loss": 3.9552,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1062876262068726e-05,
+      "loss": 3.9488,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1054506693361935e-05,
+      "loss": 3.9656,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1046120745851415e-05,
+      "loss": 3.949,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1037734798340895e-05,
+      "loss": 3.95,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1029348850830375e-05,
+      "loss": 3.9589,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1020979282123584e-05,
+      "loss": 3.9394,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101260971341679e-05,
+      "loss": 3.9511,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100422376590627e-05,
+      "loss": 3.9473,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099583781839576e-05,
+      "loss": 3.9482,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098745187088524e-05,
+      "loss": 3.955,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097906592337472e-05,
+      "loss": 3.9593,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09706799758642e-05,
+      "loss": 3.9481,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.096231040715741e-05,
+      "loss": 3.9594,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.095394083845062e-05,
+      "loss": 3.9558,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09455548909401e-05,
+      "loss": 3.9317,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093716894342958e-05,
+      "loss": 3.9479,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092879937472279e-05,
+      "loss": 3.9456,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0920413427212267e-05,
+      "loss": 3.9332,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0912027479701747e-05,
+      "loss": 3.9541,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0903641532191227e-05,
+      "loss": 3.9538,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089525558468071e-05,
+      "loss": 3.9514,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088686963717019e-05,
+      "loss": 3.942,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087848368965967e-05,
+      "loss": 3.9573,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087009774214915e-05,
+      "loss": 3.9449,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086171179463863e-05,
+      "loss": 3.9623,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085332584712811e-05,
+      "loss": 3.9347,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0844939899617586e-05,
+      "loss": 3.9485,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0836553952107066e-05,
+      "loss": 3.9562,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082818438340028e-05,
+      "loss": 3.9449,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081979843588976e-05,
+      "loss": 3.9524,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0811412488379235e-05,
+      "loss": 3.9328,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0803026540868715e-05,
+      "loss": 3.9508,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0794640593358195e-05,
+      "loss": 3.9333,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.078625464584768e-05,
+      "loss": 3.9601,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077788507714089e-05,
+      "loss": 3.9399,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076949912963037e-05,
+      "loss": 3.9512,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076111318211985e-05,
+      "loss": 3.9505,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075272723460933e-05,
+      "loss": 3.9402,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074434128709881e-05,
+      "loss": 3.9427,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073597171839202e-05,
+      "loss": 3.948,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07275857708815e-05,
+      "loss": 3.9317,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071919982337098e-05,
+      "loss": 3.9318,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071081387586046e-05,
+      "loss": 3.9527,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070242792834994e-05,
+      "loss": 3.9466,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069404198083942e-05,
+      "loss": 3.9469,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0685672412132635e-05,
+      "loss": 3.9316,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0677286464622115e-05,
+      "loss": 3.941,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0668900517111595e-05,
+      "loss": 3.9443,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0660514569601075e-05,
+      "loss": 3.9484,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0652128622090555e-05,
+      "loss": 3.9519,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0643742674580035e-05,
+      "loss": 3.9479,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0635356727069515e-05,
+      "loss": 3.9537,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0626970779558995e-05,
+      "loss": 3.943,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0618584832048475e-05,
+      "loss": 3.9397,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061023164214541e-05,
+      "loss": 3.9441,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060184569463489e-05,
+      "loss": 3.9436,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059345974712437e-05,
+      "loss": 3.9472,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058507379961385e-05,
+      "loss": 3.9423,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057668785210333e-05,
+      "loss": 3.9423,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056830190459281e-05,
+      "loss": 3.9603,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05599159570823e-05,
+      "loss": 3.9411,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055153000957178e-05,
+      "loss": 3.934,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054316044086499e-05,
+      "loss": 3.9418,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05347908721582e-05,
+      "loss": 3.947,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052640492464768e-05,
+      "loss": 3.9521,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051801897713716e-05,
+      "loss": 3.9428,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509633029626638e-05,
+      "loss": 3.9403,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0501247082116118e-05,
+      "loss": 3.9502,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0492861134605598e-05,
+      "loss": 3.9433,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0484475187095074e-05,
+      "loss": 3.9377,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0476089239584554e-05,
+      "loss": 3.944,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.046771967087777e-05,
+      "loss": 3.9469,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.045935010217098e-05,
+      "loss": 3.9534,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.045096415466046e-05,
+      "loss": 3.9474,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442578207149943e-05,
+      "loss": 3.9361,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.043420863844315e-05,
+      "loss": 3.9384,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.042582269093263e-05,
+      "loss": 3.9446,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041743674342211e-05,
+      "loss": 3.9535,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040905079591159e-05,
+      "loss": 3.9383,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040066484840107e-05,
+      "loss": 3.9395,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0392278900890548e-05,
+      "loss": 3.9363,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0383892953380028e-05,
+      "loss": 3.9426,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375507005869508e-05,
+      "loss": 3.9256,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0367137437162717e-05,
+      "loss": 3.9408,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0358751489652197e-05,
+      "loss": 3.9391,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035036554214168e-05,
+      "loss": 3.9448,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.034197959463116e-05,
+      "loss": 3.9383,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033361002592437e-05,
+      "loss": 3.9439,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032522407841385e-05,
+      "loss": 3.9308,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.031683813090333e-05,
+      "loss": 3.9427,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.030845218339281e-05,
+      "loss": 3.9454,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0300082614686022e-05,
+      "loss": 3.9333,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0291696667175502e-05,
+      "loss": 3.9472,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.028331071966498e-05,
+      "loss": 3.9516,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.027492477215446e-05,
+      "loss": 3.9457,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.026653882464394e-05,
+      "loss": 3.938,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.025815287713342e-05,
+      "loss": 3.9399,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.02497669296229e-05,
+      "loss": 3.9448,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.024138098211238e-05,
+      "loss": 3.9366,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.023299503460186e-05,
+      "loss": 3.9495,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0224609087091345e-05,
+      "loss": 3.9367,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0216223139580825e-05,
+      "loss": 3.949,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0207837192070305e-05,
+      "loss": 3.9335,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0199467623363514e-05,
+      "loss": 3.9384,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0191098054656726e-05,
+      "loss": 3.9315,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182712107146206e-05,
+      "loss": 3.9351,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174342538439415e-05,
+      "loss": 3.9276,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0165956590928895e-05,
+      "loss": 3.9497,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0157570643418375e-05,
+      "loss": 3.9583,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0149184695907855e-05,
+      "loss": 3.9382,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0140815127201068e-05,
+      "loss": 3.9341,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0132429179690548e-05,
+      "loss": 3.9475,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0124043232180028e-05,
+      "loss": 3.947,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0115657284669508e-05,
+      "loss": 3.9411,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0107271337158987e-05,
+      "loss": 3.9323,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0098885389648467e-05,
+      "loss": 3.9505,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0090499442137947e-05,
+      "loss": 3.9429,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0082113494627427e-05,
+      "loss": 3.9396,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0073727547116907e-05,
+      "loss": 3.9373,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0065341599606384e-05,
+      "loss": 3.9485,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.00569720308996e-05,
+      "loss": 3.9361,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004858608338908e-05,
+      "loss": 3.9374,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0040200135878553e-05,
+      "loss": 3.9461,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0031814188368036e-05,
+      "loss": 3.9386,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023428240857516e-05,
+      "loss": 3.9419,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0015042293346996e-05,
+      "loss": 3.9476,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0006672724640205e-05,
+      "loss": 3.9406,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.032559871673584,
+      "eval_runtime": 300.757,
+      "eval_samples_per_second": 1268.769,
+      "eval_steps_per_second": 39.65,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9998286777129685e-05,
+      "loss": 3.9296,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9989900829619165e-05,
+      "loss": 3.9355,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9981514882108645e-05,
+      "loss": 3.9502,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.997312893459813e-05,
+      "loss": 3.9463,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.996474298708761e-05,
+      "loss": 3.9475,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.995635703957709e-05,
+      "loss": 3.935,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9947971092066568e-05,
+      "loss": 3.945,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9939601523359777e-05,
+      "loss": 3.9268,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.993123195465299e-05,
+      "loss": 3.9468,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.992284600714247e-05,
+      "loss": 3.9311,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.991446005963195e-05,
+      "loss": 3.9518,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.990607411212143e-05,
+      "loss": 3.9502,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989770454341464e-05,
+      "loss": 3.9348,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.988931859590412e-05,
+      "loss": 3.9322,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.98809326483936e-05,
+      "loss": 3.9247,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9872546700883082e-05,
+      "loss": 3.9302,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9864160753372562e-05,
+      "loss": 3.9441,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9855774805862042e-05,
+      "loss": 3.9389,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9847388858351522e-05,
+      "loss": 3.9382,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.983901928964473e-05,
+      "loss": 3.9621,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.983063334213421e-05,
+      "loss": 3.946,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.982224739462369e-05,
+      "loss": 3.9467,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9813861447113174e-05,
+      "loss": 3.9358,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9805475499602654e-05,
+      "loss": 3.9544,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9797089552092134e-05,
+      "loss": 3.9335,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9788703604581614e-05,
+      "loss": 3.9453,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9780317657071094e-05,
+      "loss": 3.9436,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9771948088364303e-05,
+      "loss": 3.9337,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9763578519657516e-05,
+      "loss": 3.938,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9755192572146996e-05,
+      "loss": 3.9355,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9746806624636476e-05,
+      "loss": 3.9399,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9738420677125956e-05,
+      "loss": 3.9423,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9730051108419165e-05,
+      "loss": 3.9477,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9721665160908645e-05,
+      "loss": 3.9419,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9713295592201857e-05,
+      "loss": 3.9439,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9704909644691337e-05,
+      "loss": 3.9426,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9696523697180817e-05,
+      "loss": 3.9254,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9688137749670297e-05,
+      "loss": 3.9354,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9679751802159777e-05,
+      "loss": 3.9327,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9671365854649257e-05,
+      "loss": 3.924,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9662979907138737e-05,
+      "loss": 3.9407,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9654593959628213e-05,
+      "loss": 3.9466,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.964622439092143e-05,
+      "loss": 3.9376,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.963783844341091e-05,
+      "loss": 3.9337,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962945249590039e-05,
+      "loss": 3.9435,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9621066548389866e-05,
+      "loss": 3.9411,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9612680600879346e-05,
+      "loss": 3.9443,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9604294653368826e-05,
+      "loss": 3.9303,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9595908705858306e-05,
+      "loss": 3.9335,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9587539137151515e-05,
+      "loss": 3.944,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9579153189640995e-05,
+      "loss": 3.9375,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9570767242130475e-05,
+      "loss": 3.9376,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9562381294619958e-05,
+      "loss": 3.9252,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9553995347109438e-05,
+      "loss": 3.9405,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9545609399598918e-05,
+      "loss": 3.9227,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9537239830892127e-05,
+      "loss": 3.9484,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9528853883381607e-05,
+      "loss": 3.9299,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9520467935871087e-05,
+      "loss": 3.9407,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9512081988360567e-05,
+      "loss": 3.9352,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9503696040850047e-05,
+      "loss": 3.9348,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.949531009333953e-05,
+      "loss": 3.9274,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.948692414582901e-05,
+      "loss": 3.9391,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.947853819831849e-05,
+      "loss": 3.9223,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.947018500841543e-05,
+      "loss": 3.9212,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9461799060904908e-05,
+      "loss": 3.9412,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.945341311339439e-05,
+      "loss": 3.9382,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.944502716588387e-05,
+      "loss": 3.9351,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.943664121837335e-05,
+      "loss": 3.9182,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.942825527086283e-05,
+      "loss": 3.9284,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.941986932335231e-05,
+      "loss": 3.9347,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.941148337584179e-05,
+      "loss": 3.938,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9403113807135e-05,
+      "loss": 3.9377,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9394727859624484e-05,
+      "loss": 3.9381,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9386341912113964e-05,
+      "loss": 3.9432,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9377972343407173e-05,
+      "loss": 3.9389,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9369586395896653e-05,
+      "loss": 3.9245,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9361200448386133e-05,
+      "loss": 3.932,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9352814500875613e-05,
+      "loss": 3.9387,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9344428553365093e-05,
+      "loss": 3.9326,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9336058984658305e-05,
+      "loss": 3.933,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9327673037147785e-05,
+      "loss": 3.9323,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9319287089637265e-05,
+      "loss": 3.9501,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9310901142126745e-05,
+      "loss": 3.929,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9302515194616225e-05,
+      "loss": 3.9216,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9294145625909438e-05,
+      "loss": 3.9329,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9285759678398917e-05,
+      "loss": 3.9346,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9277373730888397e-05,
+      "loss": 3.9434,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926898778337787e-05,
+      "loss": 3.9281,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926060183586735e-05,
+      "loss": 3.933,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9252232267160566e-05,
+      "loss": 3.9366,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9243846319650046e-05,
+      "loss": 3.9338,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9235460372139523e-05,
+      "loss": 3.9243,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9227074424629003e-05,
+      "loss": 3.9328,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9218688477118483e-05,
+      "loss": 3.9404,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.92103189084117e-05,
+      "loss": 3.9479,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9201932960901175e-05,
+      "loss": 3.9317,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9193547013390655e-05,
+      "loss": 3.9244,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9185161065880135e-05,
+      "loss": 3.9316,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9176775118369615e-05,
+      "loss": 3.9308,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9168405549662824e-05,
+      "loss": 3.9443,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9160019602152304e-05,
+      "loss": 3.9312,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9151633654641784e-05,
+      "loss": 3.9287,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9143247707131268e-05,
+      "loss": 3.9228,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9134861759620747e-05,
+      "loss": 3.9372,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9126475812110227e-05,
+      "loss": 3.9082,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9118089864599707e-05,
+      "loss": 3.9315,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9109720295892916e-05,
+      "loss": 3.9261,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.910135072718613e-05,
+      "loss": 3.9367,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.909296477967561e-05,
+      "loss": 3.9262,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.908457883216509e-05,
+      "loss": 3.9342,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.907619288465457e-05,
+      "loss": 3.9235,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.906780693714405e-05,
+      "loss": 3.9314,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.905942098963353e-05,
+      "loss": 3.9343,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9051051420926738e-05,
+      "loss": 3.9238,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.904266547341622e-05,
+      "loss": 3.9393,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.90342795259057e-05,
+      "loss": 3.9432,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.902589357839518e-05,
+      "loss": 3.9351,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.901750763088466e-05,
+      "loss": 3.9286,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.900912168337414e-05,
+      "loss": 3.9296,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.900073573586362e-05,
+      "loss": 3.9322,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.89923497883531e-05,
+      "loss": 3.9297,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.898396384084258e-05,
+      "loss": 3.9362,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.897557789333206e-05,
+      "loss": 3.9291,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8967191945821537e-05,
+      "loss": 3.9367,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8958822377114753e-05,
+      "loss": 3.926,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8950436429604233e-05,
+      "loss": 3.9268,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8942050482093706e-05,
+      "loss": 3.9185,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8933680913386922e-05,
+      "loss": 3.9238,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8925294965876402e-05,
+      "loss": 3.9183,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8916909018365886e-05,
+      "loss": 3.9406,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.890852307085536e-05,
+      "loss": 3.9485,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.890013712334484e-05,
+      "loss": 3.9358,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.889175117583432e-05,
+      "loss": 3.9183,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.88833652283238e-05,
+      "loss": 3.94,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.887497928081328e-05,
+      "loss": 3.9357,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8866593333302762e-05,
+      "loss": 3.9301,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.885822376459597e-05,
+      "loss": 3.9251,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.884983781708545e-05,
+      "loss": 3.9378,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.884145186957493e-05,
+      "loss": 3.9327,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.883306592206441e-05,
+      "loss": 3.9282,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.882467997455389e-05,
+      "loss": 3.9311,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8816310405847103e-05,
+      "loss": 3.9363,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8807924458336583e-05,
+      "loss": 3.9253,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8799538510826063e-05,
+      "loss": 3.9271,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8791152563315543e-05,
+      "loss": 3.9397,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8782766615805023e-05,
+      "loss": 3.9284,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8774397047098232e-05,
+      "loss": 3.9313,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8766027478391445e-05,
+      "loss": 3.9402,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8757641530880925e-05,
+      "loss": 3.927,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0274529457092285,
+      "eval_runtime": 302.8623,
+      "eval_samples_per_second": 1259.949,
+      "eval_steps_per_second": 39.374,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8749255583370405e-05,
+      "loss": 3.9187,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8740869635859885e-05,
+      "loss": 3.9262,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8732483688349365e-05,
+      "loss": 3.9383,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8724097740838845e-05,
+      "loss": 3.9334,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8715711793328324e-05,
+      "loss": 3.9398,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8707325845817808e-05,
+      "loss": 3.9234,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8698939898307288e-05,
+      "loss": 3.934,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8690553950796768e-05,
+      "loss": 3.9167,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8682168003286248e-05,
+      "loss": 3.9372,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.867378205577572e-05,
+      "loss": 3.9244,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.86653961082652e-05,
+      "loss": 3.9389,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8657010160754684e-05,
+      "loss": 3.9344,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8648640592047893e-05,
+      "loss": 3.9286,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8640254644537373e-05,
+      "loss": 3.9221,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8631868697026853e-05,
+      "loss": 3.9192,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8623482749516333e-05,
+      "loss": 3.9203,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8615096802005813e-05,
+      "loss": 3.9319,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8606710854495293e-05,
+      "loss": 3.9304,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8598324906984776e-05,
+      "loss": 3.9266,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8589938959474256e-05,
+      "loss": 3.9502,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8581553011963736e-05,
+      "loss": 3.9403,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8573183443256945e-05,
+      "loss": 3.935,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8564797495746425e-05,
+      "loss": 3.9258,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8556411548235905e-05,
+      "loss": 3.9445,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8548025600725385e-05,
+      "loss": 3.9209,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8539639653214865e-05,
+      "loss": 3.9355,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.853125370570435e-05,
+      "loss": 3.9341,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.852286775819383e-05,
+      "loss": 3.9296,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.851448181068331e-05,
+      "loss": 3.9218,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.850609586317278e-05,
+      "loss": 3.9254,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8497726294465998e-05,
+      "loss": 3.9304,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.848935672575921e-05,
+      "loss": 3.9296,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.848097077824869e-05,
+      "loss": 3.9365,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.847258483073817e-05,
+      "loss": 3.9346,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.846419888322765e-05,
+      "loss": 3.9348,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.845581293571713e-05,
+      "loss": 3.9295,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.844744336701034e-05,
+      "loss": 3.9164,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843905741949982e-05,
+      "loss": 3.9298,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8430671471989302e-05,
+      "loss": 3.9207,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8422285524478782e-05,
+      "loss": 3.9176,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8413899576968255e-05,
+      "loss": 3.9312,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8405513629457735e-05,
+      "loss": 3.9347,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8397127681947215e-05,
+      "loss": 3.9249,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8388741734436695e-05,
+      "loss": 3.9254,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8380372165729908e-05,
+      "loss": 3.9307,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8371986218219388e-05,
+      "loss": 3.9319,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8363600270708868e-05,
+      "loss": 3.9355,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8355230702002083e-05,
+      "loss": 3.9242,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8346844754491557e-05,
+      "loss": 3.9253,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.833845880698104e-05,
+      "loss": 3.9318,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.833007285947052e-05,
+      "loss": 3.9245,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.832168691196e-05,
+      "loss": 3.9241,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.831330096444948e-05,
+      "loss": 3.9207,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.830491501693896e-05,
+      "loss": 3.9327,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.829652906942844e-05,
+      "loss": 3.9107,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.828815950072165e-05,
+      "loss": 3.9337,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8279773553211132e-05,
+      "loss": 3.9228,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8271387605700612e-05,
+      "loss": 3.9274,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8263001658190092e-05,
+      "loss": 3.9324,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8254615710679572e-05,
+      "loss": 3.9225,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8246229763169052e-05,
+      "loss": 3.9168,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8237843815658532e-05,
+      "loss": 3.9351,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8229457868148012e-05,
+      "loss": 3.9102,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8221088299441224e-05,
+      "loss": 3.9132,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8212702351930704e-05,
+      "loss": 3.9274,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8204316404420184e-05,
+      "loss": 3.9292,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8195930456909664e-05,
+      "loss": 3.9246,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8187560888202873e-05,
+      "loss": 3.9112,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8179174940692353e-05,
+      "loss": 3.9178,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8170788993181833e-05,
+      "loss": 3.9208,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8162403045671317e-05,
+      "loss": 3.9329,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8154033476964526e-05,
+      "loss": 3.9264,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8145647529454006e-05,
+      "loss": 3.9309,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8137261581943486e-05,
+      "loss": 3.9315,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8128875634432966e-05,
+      "loss": 3.9302,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8120489686922442e-05,
+      "loss": 3.9151,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8112103739411922e-05,
+      "loss": 3.921,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8103717791901402e-05,
+      "loss": 3.9329,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8095331844390882e-05,
+      "loss": 3.9196,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.808696227568409e-05,
+      "loss": 3.9253,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8078592706977307e-05,
+      "loss": 3.9187,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8070206759466787e-05,
+      "loss": 3.941,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.806182081195627e-05,
+      "loss": 3.9196,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8053434864445744e-05,
+      "loss": 3.9115,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.804506529573896e-05,
+      "loss": 3.9241,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.803667934822844e-05,
+      "loss": 3.9267,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8028293400717916e-05,
+      "loss": 3.9301,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8019907453207396e-05,
+      "loss": 3.9201,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8011537884500612e-05,
+      "loss": 3.9212,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.800315193699009e-05,
+      "loss": 3.9313,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7994765989479565e-05,
+      "loss": 3.92,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7986380041969045e-05,
+      "loss": 3.9171,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7977994094458525e-05,
+      "loss": 3.9216,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7969608146948008e-05,
+      "loss": 3.9331,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7961222199437488e-05,
+      "loss": 3.9331,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7952836251926968e-05,
+      "loss": 3.9276,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7944450304416448e-05,
+      "loss": 3.9137,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7936080735709657e-05,
+      "loss": 3.9218,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7927694788199137e-05,
+      "loss": 3.9229,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7919308840688617e-05,
+      "loss": 3.9292,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7910922893178097e-05,
+      "loss": 3.9222,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.790255332447131e-05,
+      "loss": 3.9199,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.789416737696079e-05,
+      "loss": 3.9152,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.788578142945027e-05,
+      "loss": 3.9282,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.787739548193975e-05,
+      "loss": 3.8969,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.786900953442923e-05,
+      "loss": 3.9224,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.786065634452617e-05,
+      "loss": 3.9143,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7852286775819387e-05,
+      "loss": 3.9302,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.784390082830886e-05,
+      "loss": 3.9141,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.783551488079834e-05,
+      "loss": 3.9257,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7827128933287823e-05,
+      "loss": 3.9142,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7818742985777303e-05,
+      "loss": 3.9184,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7810357038266783e-05,
+      "loss": 3.9256,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7801971090756263e-05,
+      "loss": 3.9151,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7793585143245743e-05,
+      "loss": 3.928,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7785199195735223e-05,
+      "loss": 3.9326,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7776813248224703e-05,
+      "loss": 3.9217,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7768427300714183e-05,
+      "loss": 3.9162,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7760041353203663e-05,
+      "loss": 3.9222,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7751655405693143e-05,
+      "loss": 3.9235,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7743269458182626e-05,
+      "loss": 3.9104,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7734883510672106e-05,
+      "loss": 3.9323,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.772649756316158e-05,
+      "loss": 3.9171,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.771811161565106e-05,
+      "loss": 3.9267,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7709742046944275e-05,
+      "loss": 3.9202,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7701356099433752e-05,
+      "loss": 3.9172,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.769297015192323e-05,
+      "loss": 3.9072,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.768458420441271e-05,
+      "loss": 3.9139,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7676231014509657e-05,
+      "loss": 3.9096,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7667845066999137e-05,
+      "loss": 3.929,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7659459119488617e-05,
+      "loss": 3.94,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7651073171978097e-05,
+      "loss": 3.9267,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.764268722446758e-05,
+      "loss": 3.9087,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7634301276957053e-05,
+      "loss": 3.9281,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.762593170825027e-05,
+      "loss": 3.9212,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.761754576073975e-05,
+      "loss": 3.9222,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7609159813229225e-05,
+      "loss": 3.9186,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7600773865718705e-05,
+      "loss": 3.9236,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7592387918208185e-05,
+      "loss": 3.9193,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7584001970697665e-05,
+      "loss": 3.9195,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7575632401990874e-05,
+      "loss": 3.9194,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7567246454480354e-05,
+      "loss": 3.9257,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7558860506969834e-05,
+      "loss": 3.9198,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7550474559459318e-05,
+      "loss": 3.9137,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7542088611948798e-05,
+      "loss": 3.9286,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7533719043242007e-05,
+      "loss": 3.923,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7525333095731487e-05,
+      "loss": 3.9162,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7516947148220967e-05,
+      "loss": 3.9316,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7508561200710447e-05,
+      "loss": 3.9166,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.022643566131592,
+      "eval_runtime": 292.3371,
+      "eval_samples_per_second": 1305.312,
+      "eval_steps_per_second": 40.792,
+      "step": 1373760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.750019163200366e-05,
+      "loss": 3.9131,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.749180568449314e-05,
+      "loss": 3.9149,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.748341973698262e-05,
+      "loss": 3.9301,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.74750337894721e-05,
+      "loss": 3.9249,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.746664784196158e-05,
+      "loss": 3.9274,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7458278273254788e-05,
+      "loss": 3.9138,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7449908704548e-05,
+      "loss": 3.9259,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.744152275703748e-05,
+      "loss": 3.9047,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.743313680952696e-05,
+      "loss": 3.93,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.742475086201644e-05,
+      "loss": 3.9099,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.741636491450592e-05,
+      "loss": 3.9287,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.74079789669954e-05,
+      "loss": 3.9239,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7399609398288613e-05,
+      "loss": 3.9195,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7391223450778093e-05,
+      "loss": 3.9114,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7382837503267573e-05,
+      "loss": 3.9109,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7374451555757053e-05,
+      "loss": 3.9112,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.736608198705026e-05,
+      "loss": 3.9191,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.735769603953974e-05,
+      "loss": 3.9194,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7349310092029225e-05,
+      "loss": 3.921,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7340924144518705e-05,
+      "loss": 3.9362,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7332538197008185e-05,
+      "loss": 3.9319,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7324152249497665e-05,
+      "loss": 3.927,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7315766301987145e-05,
+      "loss": 3.9185,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7307380354476625e-05,
+      "loss": 3.9285,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7298994406966105e-05,
+      "loss": 3.9153,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7290624838259314e-05,
+      "loss": 3.9238,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7282238890748797e-05,
+      "loss": 3.9233,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7273852943238277e-05,
+      "loss": 3.9186,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7265466995727757e-05,
+      "loss": 3.9108,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7257081048217237e-05,
+      "loss": 3.9183,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.724869510070671e-05,
+      "loss": 3.9206,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.724030915319619e-05,
+      "loss": 3.924,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7231923205685673e-05,
+      "loss": 3.9275,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7223553636978883e-05,
+      "loss": 3.9242,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.72151840682721e-05,
+      "loss": 3.9226,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.720679812076158e-05,
+      "loss": 3.9195,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.719841217325106e-05,
+      "loss": 3.909,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7190026225740535e-05,
+      "loss": 3.9217,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7181640278230015e-05,
+      "loss": 3.9108,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7173254330719495e-05,
+      "loss": 3.9106,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7164868383208975e-05,
+      "loss": 3.9169,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7156482435698455e-05,
+      "loss": 3.9283,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7148096488187935e-05,
+      "loss": 3.9155,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7139710540677415e-05,
+      "loss": 3.9137,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7131324593166895e-05,
+      "loss": 3.926,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7122938645656375e-05,
+      "loss": 3.9173,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7114552698145858e-05,
+      "loss": 3.9301,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7106183129439067e-05,
+      "loss": 3.9121,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7097813560732276e-05,
+      "loss": 3.9105,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7089427613221756e-05,
+      "loss": 3.9263,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7081041665711236e-05,
+      "loss": 3.9166,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.707265571820072e-05,
+      "loss": 3.9188,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.70642697706902e-05,
+      "loss": 3.9108,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.705588382317968e-05,
+      "loss": 3.9176,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.704749787566916e-05,
+      "loss": 3.9079,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.703912830696237e-05,
+      "loss": 3.9208,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7030742359451848e-05,
+      "loss": 3.9136,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7022356411941328e-05,
+      "loss": 3.9179,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.701397046443081e-05,
+      "loss": 3.9244,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.700560089572402e-05,
+      "loss": 3.9101,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.69972149482135e-05,
+      "loss": 3.907,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698882900070298e-05,
+      "loss": 3.9298,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698044305319246e-05,
+      "loss": 3.9011,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.697205710568194e-05,
+      "loss": 3.9022,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.696367115817142e-05,
+      "loss": 3.9156,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6955285210660897e-05,
+      "loss": 3.9244,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6946915641954113e-05,
+      "loss": 3.9156,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6938529694443593e-05,
+      "loss": 3.9033,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6930143746933066e-05,
+      "loss": 3.9081,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.692175779942255e-05,
+      "loss": 3.9076,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.691337185191203e-05,
+      "loss": 3.9238,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6905002283205245e-05,
+      "loss": 3.9167,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.689661633569472e-05,
+      "loss": 3.9194,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.68882303881842e-05,
+      "loss": 3.9189,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.687984444067368e-05,
+      "loss": 3.9231,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6871458493163158e-05,
+      "loss": 3.9075,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.686308892445637e-05,
+      "loss": 3.9114,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.685470297694585e-05,
+      "loss": 3.9234,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.684631702943533e-05,
+      "loss": 3.9074,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.683793108192481e-05,
+      "loss": 3.9179,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.682954513441429e-05,
+      "loss": 3.9097,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.682115918690377e-05,
+      "loss": 3.9275,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.681277323939325e-05,
+      "loss": 3.913,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.680438729188273e-05,
+      "loss": 3.9054,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6796034101979672e-05,
+      "loss": 3.91,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6787648154469152e-05,
+      "loss": 3.9171,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6779262206958632e-05,
+      "loss": 3.9262,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6770876259448112e-05,
+      "loss": 3.9109,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6762490311937595e-05,
+      "loss": 3.9135,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6754120743230804e-05,
+      "loss": 3.9105,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6745734795720284e-05,
+      "loss": 3.9197,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6737348848209764e-05,
+      "loss": 3.9096,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6728962900699244e-05,
+      "loss": 3.9081,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6720576953188724e-05,
+      "loss": 3.9238,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6712191005678204e-05,
+      "loss": 3.9226,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6703805058167684e-05,
+      "loss": 3.9206,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6695419110657167e-05,
+      "loss": 3.9059,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6687065920754106e-05,
+      "loss": 3.9127,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6678679973243586e-05,
+      "loss": 3.9146,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6670294025733066e-05,
+      "loss": 3.9193,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6661908078222546e-05,
+      "loss": 3.9167,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.665352213071203e-05,
+      "loss": 3.9086,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6645152562005238e-05,
+      "loss": 3.9067,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6636766614494718e-05,
+      "loss": 3.9197,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6628380666984198e-05,
+      "loss": 3.8919,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6619994719473678e-05,
+      "loss": 3.9106,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.661162515076689e-05,
+      "loss": 3.9041,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.66032555820601e-05,
+      "loss": 3.9211,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.659486963454958e-05,
+      "loss": 3.9097,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.658648368703906e-05,
+      "loss": 3.9135,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.657809773952854e-05,
+      "loss": 3.9055,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.656971179201802e-05,
+      "loss": 3.9092,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.65613258445075e-05,
+      "loss": 3.9187,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6552939896996983e-05,
+      "loss": 3.9089,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.654457032829019e-05,
+      "loss": 3.9183,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.65362007595834e-05,
+      "loss": 3.9216,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.652781481207288e-05,
+      "loss": 3.9154,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651942886456236e-05,
+      "loss": 3.9101,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6511042917051844e-05,
+      "loss": 3.9078,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6502656969541324e-05,
+      "loss": 3.9202,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6494271022030804e-05,
+      "loss": 3.9047,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6485885074520284e-05,
+      "loss": 3.9212,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6477499127009764e-05,
+      "loss": 3.9085,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6469113179499244e-05,
+      "loss": 3.9182,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6460727231988724e-05,
+      "loss": 3.9114,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6452341284478204e-05,
+      "loss": 3.9092,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.644395533696768e-05,
+      "loss": 3.8956,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6435602147064625e-05,
+      "loss": 3.909,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6427216199554105e-05,
+      "loss": 3.9025,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6418830252043585e-05,
+      "loss": 3.9146,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6410444304533065e-05,
+      "loss": 3.9299,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6402058357022545e-05,
+      "loss": 3.9234,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.639367240951203e-05,
+      "loss": 3.8944,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.63852864620015e-05,
+      "loss": 3.9234,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.637690051449098e-05,
+      "loss": 3.917,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636851456698046e-05,
+      "loss": 3.9135,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636012861946994e-05,
+      "loss": 3.9039,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.635174267195942e-05,
+      "loss": 3.918,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6343356724448905e-05,
+      "loss": 3.9093,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6334987155742114e-05,
+      "loss": 3.9116,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6326601208231594e-05,
+      "loss": 3.9136,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6318215260721074e-05,
+      "loss": 3.9177,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6309845692014283e-05,
+      "loss": 3.9126,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6301459744503766e-05,
+      "loss": 3.9083,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6293073796993246e-05,
+      "loss": 3.9181,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6284687849482726e-05,
+      "loss": 3.9151,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6276301901972206e-05,
+      "loss": 3.9093,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6267915954461686e-05,
+      "loss": 3.9217,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6259530006951166e-05,
+      "loss": 3.9069,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.019023895263672,
+      "eval_runtime": 291.7449,
+      "eval_samples_per_second": 1307.961,
+      "eval_steps_per_second": 40.875,
+      "step": 1450080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6251144059440646e-05,
+      "loss": 3.9059,
+      "step": 1450496
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.624277449073386e-05,
+      "loss": 3.9089,
+      "step": 1451008
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6234404922027068e-05,
+      "loss": 3.9209,
+      "step": 1451520
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6226018974516548e-05,
+      "loss": 3.9129,
+      "step": 1452032
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6217633027006027e-05,
+      "loss": 3.9272,
+      "step": 1452544
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6209247079495507e-05,
+      "loss": 3.9034,
+      "step": 1453056
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6200861131984987e-05,
+      "loss": 3.9188,
+      "step": 1453568
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.61924915632782e-05,
+      "loss": 3.8937,
+      "step": 1454080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.618412199457141e-05,
+      "loss": 3.9233,
+      "step": 1454592
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.617573604706089e-05,
+      "loss": 3.9058,
+      "step": 1455104
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.616735009955037e-05,
+      "loss": 3.9191,
+      "step": 1455616
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.615896415203985e-05,
+      "loss": 3.9162,
+      "step": 1456128
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.615059458333306e-05,
+      "loss": 3.9069,
+      "step": 1456640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.614220863582254e-05,
+      "loss": 3.9088,
+      "step": 1457152
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.613382268831202e-05,
+      "loss": 3.9045,
+      "step": 1457664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.61254367408015e-05,
+      "loss": 3.9029,
+      "step": 1458176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.611705079329098e-05,
+      "loss": 3.9141,
+      "step": 1458688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.610866484578046e-05,
+      "loss": 3.9086,
+      "step": 1459200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.610027889826994e-05,
+      "loss": 3.9133,
+      "step": 1459712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.609189295075942e-05,
+      "loss": 3.9263,
+      "step": 1460224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.60835070032489e-05,
+      "loss": 3.926,
+      "step": 1460736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6075137434542113e-05,
+      "loss": 3.9162,
+      "step": 1461248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6066751487031593e-05,
+      "loss": 3.9103,
+      "step": 1461760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6058365539521073e-05,
+      "loss": 3.9196,
+      "step": 1462272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6049979592010553e-05,
+      "loss": 3.9059,
+      "step": 1462784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6041593644500033e-05,
+      "loss": 3.9211,
+      "step": 1463296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.603320769698951e-05,
+      "loss": 3.9148,
+      "step": 1463808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.602482174947899e-05,
+      "loss": 3.9083,
+      "step": 1464320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.601643580196847e-05,
+      "loss": 3.902,
+      "step": 1464832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.600804985445795e-05,
+      "loss": 3.9128,
+      "step": 1465344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.599966390694743e-05,
+      "loss": 3.9075,
+      "step": 1465856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.599127795943691e-05,
+      "loss": 3.9168,
+      "step": 1466368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5982908390730122e-05,
+      "loss": 3.9185,
+      "step": 1466880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5974522443219602e-05,
+      "loss": 3.9136,
+      "step": 1467392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.596615287451281e-05,
+      "loss": 3.9147,
+      "step": 1467904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.595776692700229e-05,
+      "loss": 3.9149,
+      "step": 1468416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.594938097949177e-05,
+      "loss": 3.899,
+      "step": 1468928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.594099503198125e-05,
+      "loss": 3.9108,
+      "step": 1469440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.593260908447073e-05,
+      "loss": 3.8994,
+      "step": 1469952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5924223136960214e-05,
+      "loss": 3.9066,
+      "step": 1470464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5915837189449694e-05,
+      "loss": 3.909,
+      "step": 1470976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5907451241939174e-05,
+      "loss": 3.9163,
+      "step": 1471488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5899065294428654e-05,
+      "loss": 3.9131,
+      "step": 1472000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5890679346918134e-05,
+      "loss": 3.9021,
+      "step": 1472512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5882293399407614e-05,
+      "loss": 3.9162,
+      "step": 1473024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5873907451897094e-05,
+      "loss": 3.9147,
+      "step": 1473536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5865554261994036e-05,
+      "loss": 3.9189,
+      "step": 1474048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5857168314483516e-05,
+      "loss": 3.9093,
+      "step": 1474560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5848782366972996e-05,
+      "loss": 3.9005,
+      "step": 1475072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5840396419462476e-05,
+      "loss": 3.9185,
+      "step": 1475584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5832010471951955e-05,
+      "loss": 3.9076,
+      "step": 1476096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5823624524441435e-05,
+      "loss": 3.9099,
+      "step": 1476608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5815238576930915e-05,
+      "loss": 3.9058,
+      "step": 1477120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.58068526294204e-05,
+      "loss": 3.9058,
+      "step": 1477632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5798483060713608e-05,
+      "loss": 3.9021,
+      "step": 1478144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5790097113203088e-05,
+      "loss": 3.9086,
+      "step": 1478656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5781711165692568e-05,
+      "loss": 3.9087,
+      "step": 1479168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5773325218182048e-05,
+      "loss": 3.9121,
+      "step": 1479680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5764939270671524e-05,
+      "loss": 3.9115,
+      "step": 1480192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5756553323161004e-05,
+      "loss": 3.901,
+      "step": 1480704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5748167375650484e-05,
+      "loss": 3.9012,
+      "step": 1481216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5739781428139964e-05,
+      "loss": 3.9237,
+      "step": 1481728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5731411859433173e-05,
+      "loss": 3.894,
+      "step": 1482240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5723025911922653e-05,
+      "loss": 3.8941,
+      "step": 1482752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5714639964412137e-05,
+      "loss": 3.9054,
+      "step": 1483264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5706254016901616e-05,
+      "loss": 3.9189,
+      "step": 1483776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5697868069391096e-05,
+      "loss": 3.9049,
+      "step": 1484288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5689498500684306e-05,
+      "loss": 3.8993,
+      "step": 1484800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5681112553173786e-05,
+      "loss": 3.8958,
+      "step": 1485312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5672742984466998e-05,
+      "loss": 3.8989,
+      "step": 1485824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5664357036956478e-05,
+      "loss": 3.9155,
+      "step": 1486336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5655971089445958e-05,
+      "loss": 3.9068,
+      "step": 1486848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5647585141935438e-05,
+      "loss": 3.9078,
+      "step": 1487360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5639199194424918e-05,
+      "loss": 3.9169,
+      "step": 1487872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5630829625718127e-05,
+      "loss": 3.9165,
+      "step": 1488384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5622443678207607e-05,
+      "loss": 3.8995,
+      "step": 1488896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.561407410950082e-05,
+      "loss": 3.9047,
+      "step": 1489408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.56056881619903e-05,
+      "loss": 3.9176,
+      "step": 1489920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.559730221447978e-05,
+      "loss": 3.8946,
+      "step": 1490432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.558891626696926e-05,
+      "loss": 3.9121,
+      "step": 1490944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.558053031945874e-05,
+      "loss": 3.9003,
+      "step": 1491456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.557214437194822e-05,
+      "loss": 3.923,
+      "step": 1491968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.55637584244377e-05,
+      "loss": 3.905,
+      "step": 1492480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.555537247692718e-05,
+      "loss": 3.8993,
+      "step": 1492992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.554700290822039e-05,
+      "loss": 3.8984,
+      "step": 1493504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.553861696070987e-05,
+      "loss": 3.9103,
+      "step": 1494016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.553024739200308e-05,
+      "loss": 3.9173,
+      "step": 1494528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.552186144449256e-05,
+      "loss": 3.9062,
+      "step": 1495040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5513491875785773e-05,
+      "loss": 3.9036,
+      "step": 1495552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5505105928275253e-05,
+      "loss": 3.9017,
+      "step": 1496064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5496719980764733e-05,
+      "loss": 3.9122,
+      "step": 1496576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5488334033254213e-05,
+      "loss": 3.9033,
+      "step": 1497088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5479948085743693e-05,
+      "loss": 3.9015,
+      "step": 1497600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5471562138233173e-05,
+      "loss": 3.9153,
+      "step": 1498112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5463176190722653e-05,
+      "loss": 3.9157,
+      "step": 1498624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5454790243212133e-05,
+      "loss": 3.9103,
+      "step": 1499136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5446404295701616e-05,
+      "loss": 3.8946,
+      "step": 1499648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5438018348191096e-05,
+      "loss": 3.9085,
+      "step": 1500160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5429632400680576e-05,
+      "loss": 3.9027,
+      "step": 1500672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5421246453170056e-05,
+      "loss": 3.9116,
+      "step": 1501184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5412876884463265e-05,
+      "loss": 3.9095,
+      "step": 1501696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5404490936952745e-05,
+      "loss": 3.9003,
+      "step": 1502208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5396104989442225e-05,
+      "loss": 3.903,
+      "step": 1502720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5387719041931708e-05,
+      "loss": 3.9096,
+      "step": 1503232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5379349473224917e-05,
+      "loss": 3.887,
+      "step": 1503744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5370963525714397e-05,
+      "loss": 3.9065,
+      "step": 1504256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5362577578203877e-05,
+      "loss": 3.8917,
+      "step": 1504768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5354208009497086e-05,
+      "loss": 3.9145,
+      "step": 1505280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.534582206198657e-05,
+      "loss": 3.9038,
+      "step": 1505792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.533743611447605e-05,
+      "loss": 3.9031,
+      "step": 1506304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.532905016696553e-05,
+      "loss": 3.8987,
+      "step": 1506816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5320664219455003e-05,
+      "loss": 3.9052,
+      "step": 1507328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.531229465074822e-05,
+      "loss": 3.9104,
+      "step": 1507840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.53039087032377e-05,
+      "loss": 3.8981,
+      "step": 1508352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.529552275572718e-05,
+      "loss": 3.9081,
+      "step": 1508864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5287136808216655e-05,
+      "loss": 3.913,
+      "step": 1509376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5278750860706135e-05,
+      "loss": 3.9067,
+      "step": 1509888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5270364913195615e-05,
+      "loss": 3.9098,
+      "step": 1510400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5261978965685095e-05,
+      "loss": 3.8961,
+      "step": 1510912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5253609396978308e-05,
+      "loss": 3.9148,
+      "step": 1511424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5245223449467787e-05,
+      "loss": 3.9002,
+      "step": 1511936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5236837501957267e-05,
+      "loss": 3.9075,
+      "step": 1512448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5228451554446747e-05,
+      "loss": 3.8993,
+      "step": 1512960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5220065606936227e-05,
+      "loss": 3.9091,
+      "step": 1513472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5211679659425707e-05,
+      "loss": 3.9061,
+      "step": 1513984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5203293711915187e-05,
+      "loss": 3.8989,
+      "step": 1514496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5194907764404667e-05,
+      "loss": 3.8925,
+      "step": 1515008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5186521816894147e-05,
+      "loss": 3.8977,
+      "step": 1515520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.517813586938363e-05,
+      "loss": 3.8999,
+      "step": 1516032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.516974992187311e-05,
+      "loss": 3.906,
+      "step": 1516544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.516136397436259e-05,
+      "loss": 3.9191,
+      "step": 1517056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.51529944056558e-05,
+      "loss": 3.9164,
+      "step": 1517568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.514462483694901e-05,
+      "loss": 3.8891,
+      "step": 1518080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5136238889438492e-05,
+      "loss": 3.9144,
+      "step": 1518592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5127852941927972e-05,
+      "loss": 3.9042,
+      "step": 1519104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.511948337322118e-05,
+      "loss": 3.9052,
+      "step": 1519616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.511109742571066e-05,
+      "loss": 3.8993,
+      "step": 1520128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.510271147820014e-05,
+      "loss": 3.9072,
+      "step": 1520640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.509432553068962e-05,
+      "loss": 3.9015,
+      "step": 1521152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.50859395831791e-05,
+      "loss": 3.9099,
+      "step": 1521664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5077553635668584e-05,
+      "loss": 3.9029,
+      "step": 1522176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5069167688158064e-05,
+      "loss": 3.9073,
+      "step": 1522688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5060781740647544e-05,
+      "loss": 3.9092,
+      "step": 1523200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5052395793137017e-05,
+      "loss": 3.8981,
+      "step": 1523712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5044026224430233e-05,
+      "loss": 3.9073,
+      "step": 1524224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5035640276919713e-05,
+      "loss": 3.9128,
+      "step": 1524736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5027270708212926e-05,
+      "loss": 3.8994,
+      "step": 1525248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5018884760702406e-05,
+      "loss": 3.9096,
+      "step": 1525760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5010498813191885e-05,
+      "loss": 3.9057,
+      "step": 1526272
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.015557289123535,
+      "eval_runtime": 290.6984,
+      "eval_samples_per_second": 1312.67,
+      "eval_steps_per_second": 41.022,
+      "step": 1526400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5002112865681365e-05,
+      "loss": 3.8975,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4993726918170842e-05,
+      "loss": 3.8994,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4985340970660322e-05,
+      "loss": 3.9083,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4976955023149802e-05,
+      "loss": 3.907,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4968569075639282e-05,
+      "loss": 3.9209,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4960183128128762e-05,
+      "loss": 3.894,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4951797180618242e-05,
+      "loss": 3.9072,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4943411233107722e-05,
+      "loss": 3.8939,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.49350252855972e-05,
+      "loss": 3.9083,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.492663933808668e-05,
+      "loss": 3.8997,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.491825339057616e-05,
+      "loss": 3.9074,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.490986744306564e-05,
+      "loss": 3.9115,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4901497874358854e-05,
+      "loss": 3.8973,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4893111926848334e-05,
+      "loss": 3.9035,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4884725979337814e-05,
+      "loss": 3.8948,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.487634003182729e-05,
+      "loss": 3.8914,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.486795408431677e-05,
+      "loss": 3.9042,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4859568136806254e-05,
+      "loss": 3.9021,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4851182189295734e-05,
+      "loss": 3.9039,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4842796241785214e-05,
+      "loss": 3.9159,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4834410294274694e-05,
+      "loss": 3.919,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4826040725567903e-05,
+      "loss": 3.9093,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4817654778057383e-05,
+      "loss": 3.9054,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4809268830546863e-05,
+      "loss": 3.9142,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4800899261840075e-05,
+      "loss": 3.8979,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4792529693133288e-05,
+      "loss": 3.912,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4784143745622764e-05,
+      "loss": 3.908,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4775757798112244e-05,
+      "loss": 3.8985,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4767371850601724e-05,
+      "loss": 3.898,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4758985903091207e-05,
+      "loss": 3.9025,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4750599955580687e-05,
+      "loss": 3.9026,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4742214008070167e-05,
+      "loss": 3.9072,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4733828060559644e-05,
+      "loss": 3.9128,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4725442113049124e-05,
+      "loss": 3.9076,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4717056165538604e-05,
+      "loss": 3.9079,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4708670218028084e-05,
+      "loss": 3.9068,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4700284270517564e-05,
+      "loss": 3.8898,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4691898323007047e-05,
+      "loss": 3.906,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4683512375496527e-05,
+      "loss": 3.8945,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4675142806789736e-05,
+      "loss": 3.8978,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4666756859279216e-05,
+      "loss": 3.899,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4658370911768696e-05,
+      "loss": 3.9136,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.465000134306191e-05,
+      "loss": 3.9037,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464161539555139e-05,
+      "loss": 3.894,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.463322944804087e-05,
+      "loss": 3.9117,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.462484350053035e-05,
+      "loss": 3.9039,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4616457553019825e-05,
+      "loss": 3.908,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4608071605509305e-05,
+      "loss": 3.9047,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4599685657998785e-05,
+      "loss": 3.8913,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4591299710488265e-05,
+      "loss": 3.9127,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4582930141781477e-05,
+      "loss": 3.8974,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4574544194270957e-05,
+      "loss": 3.9055,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4566158246760437e-05,
+      "loss": 3.8971,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557772299249917e-05,
+      "loss": 3.894,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549386351739397e-05,
+      "loss": 3.8976,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4541000404228877e-05,
+      "loss": 3.903,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4532614456718357e-05,
+      "loss": 3.8994,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4524228509207837e-05,
+      "loss": 3.9033,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.451587531930478e-05,
+      "loss": 3.9071,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.450748937179426e-05,
+      "loss": 3.8963,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.449910342428374e-05,
+      "loss": 3.8907,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.449071747677322e-05,
+      "loss": 3.9175,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4482331529262702e-05,
+      "loss": 3.884,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4473945581752182e-05,
+      "loss": 3.8858,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.446555963424166e-05,
+      "loss": 3.8971,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.445717368673114e-05,
+      "loss": 3.913,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.444880411802435e-05,
+      "loss": 3.8933,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4440418170513827e-05,
+      "loss": 3.8979,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.443203222300331e-05,
+      "loss": 3.8871,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.442364627549279e-05,
+      "loss": 3.8923,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.441526032798227e-05,
+      "loss": 3.9057,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.440687438047175e-05,
+      "loss": 3.9026,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.439848843296123e-05,
+      "loss": 3.9019,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.439011886425444e-05,
+      "loss": 3.9084,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.438173291674392e-05,
+      "loss": 3.9093,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373346969233403e-05,
+      "loss": 3.8947,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4364961021722883e-05,
+      "loss": 3.8955,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4356575074212363e-05,
+      "loss": 3.9103,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.434818912670184e-05,
+      "loss": 3.8869,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.433980317919132e-05,
+      "loss": 3.9041,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.43314172316808e-05,
+      "loss": 3.8962,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4323047662974012e-05,
+      "loss": 3.9147,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4314678094267224e-05,
+      "loss": 3.9005,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4306292146756704e-05,
+      "loss": 3.8924,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4297922578049913e-05,
+      "loss": 3.885,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4289536630539393e-05,
+      "loss": 3.9071,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4281150683028873e-05,
+      "loss": 3.9075,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4272764735518357e-05,
+      "loss": 3.9025,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4264395166811566e-05,
+      "loss": 3.8956,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4256009219301046e-05,
+      "loss": 3.8974,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4247623271790526e-05,
+      "loss": 3.9036,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4239237324280006e-05,
+      "loss": 3.8947,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4230851376769486e-05,
+      "loss": 3.8946,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4222465429258966e-05,
+      "loss": 3.9091,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4214079481748445e-05,
+      "loss": 3.9063,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205693534237925e-05,
+      "loss": 3.9013,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4197307586727405e-05,
+      "loss": 3.8916,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4188921639216885e-05,
+      "loss": 3.9004,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4180535691706365e-05,
+      "loss": 3.8947,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4172149744195842e-05,
+      "loss": 3.9041,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4163780175489058e-05,
+      "loss": 3.9006,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4155410606782267e-05,
+      "loss": 3.8996,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4147024659271747e-05,
+      "loss": 3.894,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4138638711761227e-05,
+      "loss": 3.8973,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4130252764250707e-05,
+      "loss": 3.8881,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4121866816740187e-05,
+      "loss": 3.8927,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4113480869229667e-05,
+      "loss": 3.8846,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.410511130052288e-05,
+      "loss": 3.9076,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.409672535301236e-05,
+      "loss": 3.8968,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.408833940550184e-05,
+      "loss": 3.8942,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4079953457991316e-05,
+      "loss": 3.8891,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4071567510480796e-05,
+      "loss": 3.9015,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.406318156297028e-05,
+      "loss": 3.9018,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.405479561545976e-05,
+      "loss": 3.8919,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4046426046752968e-05,
+      "loss": 3.9022,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4038040099242448e-05,
+      "loss": 3.909,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4029654151731928e-05,
+      "loss": 3.898,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4021268204221408e-05,
+      "loss": 3.9026,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4012882256710888e-05,
+      "loss": 3.889,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.400449630920037e-05,
+      "loss": 3.902,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.399612674049358e-05,
+      "loss": 3.8973,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.398774079298306e-05,
+      "loss": 3.9042,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.397935484547254e-05,
+      "loss": 3.8863,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.397096889796202e-05,
+      "loss": 3.9051,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3962582950451497e-05,
+      "loss": 3.8991,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3954213381744712e-05,
+      "loss": 3.892,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3945827434234192e-05,
+      "loss": 3.8856,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.393744148672367e-05,
+      "loss": 3.8905,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392905553921315e-05,
+      "loss": 3.8886,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392066959170263e-05,
+      "loss": 3.9022,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.391228364419211e-05,
+      "loss": 3.9076,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.390389769668159e-05,
+      "loss": 3.911,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3895511749171072e-05,
+      "loss": 3.8825,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.388714218046428e-05,
+      "loss": 3.9066,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387875623295376e-05,
+      "loss": 3.9023,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387038666424697e-05,
+      "loss": 3.8944,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.386200071673645e-05,
+      "loss": 3.8929,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3853614769225934e-05,
+      "loss": 3.9046,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3845228821715414e-05,
+      "loss": 3.8959,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3836842874204894e-05,
+      "loss": 3.8998,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3828456926694374e-05,
+      "loss": 3.8909,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382007097918385e-05,
+      "loss": 3.898,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3811701410477063e-05,
+      "loss": 3.9009,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3803315462966543e-05,
+      "loss": 3.8952,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3794929515456026e-05,
+      "loss": 3.8983,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3786543567945502e-05,
+      "loss": 3.906,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3778157620434982e-05,
+      "loss": 3.894,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3769788051728195e-05,
+      "loss": 3.9014,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3761402104217675e-05,
+      "loss": 3.9015,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.012962341308594,
+      "eval_runtime": 293.1665,
+      "eval_samples_per_second": 1301.619,
+      "eval_steps_per_second": 40.677,
+      "step": 1602720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.375301615670715e-05,
+      "loss": 3.8926,
+      "step": 1603072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3744630209196635e-05,
+      "loss": 3.8905,
+      "step": 1603584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3736244261686115e-05,
+      "loss": 3.9034,
+      "step": 1604096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3727858314175595e-05,
+      "loss": 3.9018,
+      "step": 1604608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3719472366665075e-05,
+      "loss": 3.9105,
+      "step": 1605120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3711086419154555e-05,
+      "loss": 3.8914,
+      "step": 1605632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.370270047164403e-05,
+      "loss": 3.8999,
+      "step": 1606144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.369431452413351e-05,
+      "loss": 3.8872,
+      "step": 1606656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.368592857662299e-05,
+      "loss": 3.899,
+      "step": 1607168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3677542629112474e-05,
+      "loss": 3.8962,
+      "step": 1607680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3669173060405683e-05,
+      "loss": 3.899,
+      "step": 1608192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3660787112895163e-05,
+      "loss": 3.9034,
+      "step": 1608704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3652417544188376e-05,
+      "loss": 3.8939,
+      "step": 1609216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3644031596677856e-05,
+      "loss": 3.8953,
+      "step": 1609728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3635645649167336e-05,
+      "loss": 3.891,
+      "step": 1610240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3627259701656816e-05,
+      "loss": 3.8834,
+      "step": 1610752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3618873754146296e-05,
+      "loss": 3.8933,
+      "step": 1611264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3610487806635776e-05,
+      "loss": 3.8974,
+      "step": 1611776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3602101859125256e-05,
+      "loss": 3.8954,
+      "step": 1612288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3593732290418465e-05,
+      "loss": 3.9119,
+      "step": 1612800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3585346342907945e-05,
+      "loss": 3.9112,
+      "step": 1613312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3576960395397428e-05,
+      "loss": 3.9054,
+      "step": 1613824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3568574447886908e-05,
+      "loss": 3.8972,
+      "step": 1614336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3560188500376388e-05,
+      "loss": 3.905,
+      "step": 1614848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3551818931669597e-05,
+      "loss": 3.8925,
+      "step": 1615360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3543432984159077e-05,
+      "loss": 3.9046,
+      "step": 1615872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3535047036648557e-05,
+      "loss": 3.8985,
+      "step": 1616384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3526661089138037e-05,
+      "loss": 3.8882,
+      "step": 1616896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3518275141627517e-05,
+      "loss": 3.8905,
+      "step": 1617408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.350990557292073e-05,
+      "loss": 3.8989,
+      "step": 1617920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.350151962541021e-05,
+      "loss": 3.8944,
+      "step": 1618432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3493133677899686e-05,
+      "loss": 3.8978,
+      "step": 1618944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3484747730389166e-05,
+      "loss": 3.9066,
+      "step": 1619456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.347636178287865e-05,
+      "loss": 3.9006,
+      "step": 1619968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.346799221417186e-05,
+      "loss": 3.9004,
+      "step": 1620480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3459606266661338e-05,
+      "loss": 3.9007,
+      "step": 1620992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.345123669795455e-05,
+      "loss": 3.8892,
+      "step": 1621504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.344285075044403e-05,
+      "loss": 3.8907,
+      "step": 1622016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.343446480293351e-05,
+      "loss": 3.8866,
+      "step": 1622528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.342607885542299e-05,
+      "loss": 3.8898,
+      "step": 1623040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.341769290791247e-05,
+      "loss": 3.8927,
+      "step": 1623552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3409323339205683e-05,
+      "loss": 3.9069,
+      "step": 1624064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.340093739169516e-05,
+      "loss": 3.8966,
+      "step": 1624576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.339255144418464e-05,
+      "loss": 3.8878,
+      "step": 1625088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.338416549667412e-05,
+      "loss": 3.9074,
+      "step": 1625600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.33757795491636e-05,
+      "loss": 3.8926,
+      "step": 1626112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3367393601653083e-05,
+      "loss": 3.9066,
+      "step": 1626624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3359007654142563e-05,
+      "loss": 3.8954,
+      "step": 1627136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3350621706632043e-05,
+      "loss": 3.8845,
+      "step": 1627648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.334223575912152e-05,
+      "loss": 3.903,
+      "step": 1628160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3333866190414732e-05,
+      "loss": 3.8921,
+      "step": 1628672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3325496621707944e-05,
+      "loss": 3.9014,
+      "step": 1629184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3317110674197424e-05,
+      "loss": 3.8869,
+      "step": 1629696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3308724726686904e-05,
+      "loss": 3.8906,
+      "step": 1630208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3300338779176384e-05,
+      "loss": 3.8881,
+      "step": 1630720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3291952831665864e-05,
+      "loss": 3.8946,
+      "step": 1631232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.328356688415534e-05,
+      "loss": 3.8956,
+      "step": 1631744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.327518093664482e-05,
+      "loss": 3.89,
+      "step": 1632256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3266794989134304e-05,
+      "loss": 3.9028,
+      "step": 1632768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3258409041623784e-05,
+      "loss": 3.8906,
+      "step": 1633280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3250039472916993e-05,
+      "loss": 3.8837,
+      "step": 1633792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3241653525406473e-05,
+      "loss": 3.9116,
+      "step": 1634304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3233267577895953e-05,
+      "loss": 3.8773,
+      "step": 1634816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3224881630385433e-05,
+      "loss": 3.8748,
+      "step": 1635328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3216495682874913e-05,
+      "loss": 3.8894,
+      "step": 1635840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3208109735364396e-05,
+      "loss": 3.9098,
+      "step": 1636352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3199723787853873e-05,
+      "loss": 3.8829,
+      "step": 1636864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3191354219147085e-05,
+      "loss": 3.8918,
+      "step": 1637376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3182968271636565e-05,
+      "loss": 3.8798,
+      "step": 1637888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3174582324126045e-05,
+      "loss": 3.886,
+      "step": 1638400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.316619637661552e-05,
+      "loss": 3.8984,
+      "step": 1638912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3157810429105005e-05,
+      "loss": 3.8966,
+      "step": 1639424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3149424481594485e-05,
+      "loss": 3.8948,
+      "step": 1639936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3141038534083965e-05,
+      "loss": 3.8966,
+      "step": 1640448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3132652586573445e-05,
+      "loss": 3.9034,
+      "step": 1640960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3124283017866654e-05,
+      "loss": 3.8852,
+      "step": 1641472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3115897070356134e-05,
+      "loss": 3.8934,
+      "step": 1641984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3107527501649346e-05,
+      "loss": 3.899,
+      "step": 1642496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3099141554138826e-05,
+      "loss": 3.8814,
+      "step": 1643008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3090755606628306e-05,
+      "loss": 3.8936,
+      "step": 1643520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3082369659117786e-05,
+      "loss": 3.8898,
+      "step": 1644032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3073983711607266e-05,
+      "loss": 3.9029,
+      "step": 1644544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3065597764096746e-05,
+      "loss": 3.895,
+      "step": 1645056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3057211816586226e-05,
+      "loss": 3.8856,
+      "step": 1645568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304884224787944e-05,
+      "loss": 3.8824,
+      "step": 1646080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304045630036892e-05,
+      "loss": 3.8952,
+      "step": 1646592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.30320703528584e-05,
+      "loss": 3.8982,
+      "step": 1647104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.302368440534788e-05,
+      "loss": 3.8891,
+      "step": 1647616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3015314836641088e-05,
+      "loss": 3.8942,
+      "step": 1648128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3006928889130568e-05,
+      "loss": 3.8896,
+      "step": 1648640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.299854294162005e-05,
+      "loss": 3.8953,
+      "step": 1649152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.299017337291326e-05,
+      "loss": 3.8886,
+      "step": 1649664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.298178742540274e-05,
+      "loss": 3.8853,
+      "step": 1650176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.297340147789222e-05,
+      "loss": 3.9035,
+      "step": 1650688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.29650155303817e-05,
+      "loss": 3.8978,
+      "step": 1651200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2956629582871176e-05,
+      "loss": 3.8956,
+      "step": 1651712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.294824363536066e-05,
+      "loss": 3.8828,
+      "step": 1652224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2939874066653872e-05,
+      "loss": 3.8961,
+      "step": 1652736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2931488119143352e-05,
+      "loss": 3.8858,
+      "step": 1653248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.292310217163283e-05,
+      "loss": 3.8988,
+      "step": 1653760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.291471622412231e-05,
+      "loss": 3.8942,
+      "step": 1654272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.290633027661179e-05,
+      "loss": 3.8912,
+      "step": 1654784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.289794432910127e-05,
+      "loss": 3.8882,
+      "step": 1655296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2889558381590752e-05,
+      "loss": 3.8822,
+      "step": 1655808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2881172434080232e-05,
+      "loss": 3.8867,
+      "step": 1656320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.287278648656971e-05,
+      "loss": 3.881,
+      "step": 1656832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.286440053905919e-05,
+      "loss": 3.882,
+      "step": 1657344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.28560309703524e-05,
+      "loss": 3.8963,
+      "step": 1657856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.284764502284188e-05,
+      "loss": 3.894,
+      "step": 1658368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.283925907533136e-05,
+      "loss": 3.8886,
+      "step": 1658880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.283087312782084e-05,
+      "loss": 3.8803,
+      "step": 1659392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2822503559114053e-05,
+      "loss": 3.8916,
+      "step": 1659904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2814117611603533e-05,
+      "loss": 3.8964,
+      "step": 1660416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.280573166409301e-05,
+      "loss": 3.8876,
+      "step": 1660928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.279734571658249e-05,
+      "loss": 3.8953,
+      "step": 1661440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.278895976907197e-05,
+      "loss": 3.8965,
+      "step": 1661952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2780590200365182e-05,
+      "loss": 3.8943,
+      "step": 1662464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2772204252854662e-05,
+      "loss": 3.8911,
+      "step": 1662976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2763834684147875e-05,
+      "loss": 3.885,
+      "step": 1663488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2755448736637355e-05,
+      "loss": 3.8913,
+      "step": 1664000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.274706278912683e-05,
+      "loss": 3.8908,
+      "step": 1664512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2738676841616315e-05,
+      "loss": 3.8962,
+      "step": 1665024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2730290894105795e-05,
+      "loss": 3.8827,
+      "step": 1665536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2721904946595274e-05,
+      "loss": 3.8949,
+      "step": 1666048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2713518999084754e-05,
+      "loss": 3.8949,
+      "step": 1666560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2705133051574234e-05,
+      "loss": 3.8875,
+      "step": 1667072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2696763482867443e-05,
+      "loss": 3.8759,
+      "step": 1667584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2688377535356923e-05,
+      "loss": 3.8881,
+      "step": 1668096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2679991587846407e-05,
+      "loss": 3.8813,
+      "step": 1668608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2671605640335887e-05,
+      "loss": 3.8947,
+      "step": 1669120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2663219692825363e-05,
+      "loss": 3.9005,
+      "step": 1669632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2654850124118576e-05,
+      "loss": 3.9037,
+      "step": 1670144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2646464176608056e-05,
+      "loss": 3.8766,
+      "step": 1670656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2638078229097536e-05,
+      "loss": 3.8994,
+      "step": 1671168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2629692281587016e-05,
+      "loss": 3.8962,
+      "step": 1671680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2621322712880228e-05,
+      "loss": 3.8879,
+      "step": 1672192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2612936765369708e-05,
+      "loss": 3.8835,
+      "step": 1672704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2604550817859185e-05,
+      "loss": 3.8998,
+      "step": 1673216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2596164870348665e-05,
+      "loss": 3.8843,
+      "step": 1673728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2587778922838145e-05,
+      "loss": 3.8962,
+      "step": 1674240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2579392975327628e-05,
+      "loss": 3.8819,
+      "step": 1674752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2571007027817108e-05,
+      "loss": 3.8947,
+      "step": 1675264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2562621080306588e-05,
+      "loss": 3.8925,
+      "step": 1675776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2554251511599797e-05,
+      "loss": 3.8893,
+      "step": 1676288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2545865564089277e-05,
+      "loss": 3.8958,
+      "step": 1676800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.253749599538249e-05,
+      "loss": 3.8911,
+      "step": 1677312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252911004787197e-05,
+      "loss": 3.8926,
+      "step": 1677824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252072410036145e-05,
+      "loss": 3.8869,
+      "step": 1678336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.251233815285093e-05,
+      "loss": 3.8975,
+      "step": 1678848
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.009742259979248,
+      "eval_runtime": 293.1704,
+      "eval_samples_per_second": 1301.601,
+      "eval_steps_per_second": 40.676,
+      "step": 1679040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.250395220534041e-05,
+      "loss": 3.8885,
+      "step": 1679360
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.249556625782989e-05,
+      "loss": 3.8818,
+      "step": 1679872
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2487180310319366e-05,
+      "loss": 3.8919,
+      "step": 1680384
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2478794362808846e-05,
+      "loss": 3.8971,
+      "step": 1680896
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.247040841529833e-05,
+      "loss": 3.9011,
+      "step": 1681408
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.246202246778781e-05,
+      "loss": 3.8863,
+      "step": 1681920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2453652899081018e-05,
+      "loss": 3.8898,
+      "step": 1682432
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2445266951570498e-05,
+      "loss": 3.8784,
+      "step": 1682944
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.243689738286371e-05,
+      "loss": 3.8935,
+      "step": 1683456
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.242851143535319e-05,
+      "loss": 3.8862,
+      "step": 1683968
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.242012548784267e-05,
+      "loss": 3.8958,
+      "step": 1684480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.241173954033215e-05,
+      "loss": 3.8943,
+      "step": 1684992
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2403369971625363e-05,
+      "loss": 3.8874,
+      "step": 1685504
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.239498402411484e-05,
+      "loss": 3.891,
+      "step": 1686016
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.238659807660432e-05,
+      "loss": 3.8788,
+      "step": 1686528
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.23782121290938e-05,
+      "loss": 3.8758,
+      "step": 1687040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2369826181583283e-05,
+      "loss": 3.8879,
+      "step": 1687552
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2361440234072763e-05,
+      "loss": 3.8948,
+      "step": 1688064
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2353054286562243e-05,
+      "loss": 3.8875,
+      "step": 1688576
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2344668339051723e-05,
+      "loss": 3.9063,
+      "step": 1689088
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.23362823915412e-05,
+      "loss": 3.9004,
+      "step": 1689600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.232789644403068e-05,
+      "loss": 3.8994,
+      "step": 1690112
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.231951049652016e-05,
+      "loss": 3.8897,
+      "step": 1690624
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.231112454900964e-05,
+      "loss": 3.8967,
+      "step": 1691136
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2302771359106584e-05,
+      "loss": 3.8825,
+      "step": 1691648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2294385411596064e-05,
+      "loss": 3.9002,
+      "step": 1692160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2285999464085544e-05,
+      "loss": 3.8899,
+      "step": 1692672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.227761351657502e-05,
+      "loss": 3.8836,
+      "step": 1693184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.22692275690645e-05,
+      "loss": 3.8889,
+      "step": 1693696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.2260841621553984e-05,
+      "loss": 3.8897,
+      "step": 1694208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2252455674043464e-05,
+      "loss": 3.8886,
+      "step": 1694720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2244069726532944e-05,
+      "loss": 3.8932,
+      "step": 1695232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2235683779022424e-05,
+      "loss": 3.8947,
+      "step": 1695744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2227314210315633e-05,
+      "loss": 3.8983,
+      "step": 1696256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2218928262805113e-05,
+      "loss": 3.8874,
+      "step": 1696768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2210542315294593e-05,
+      "loss": 3.8925,
+      "step": 1697280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2202156367784076e-05,
+      "loss": 3.8874,
+      "step": 1697792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2193786799077285e-05,
+      "loss": 3.8782,
+      "step": 1698304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2185400851566765e-05,
+      "loss": 3.8782,
+      "step": 1698816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2177014904056245e-05,
+      "loss": 3.8826,
+      "step": 1699328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2168628956545725e-05,
+      "loss": 3.8848,
+      "step": 1699840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.21602430090352e-05,
+      "loss": 3.8997,
+      "step": 1700352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2151857061524685e-05,
+      "loss": 3.8878,
+      "step": 1700864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2143487492817897e-05,
+      "loss": 3.8826,
+      "step": 1701376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2135101545307377e-05,
+      "loss": 3.8996,
+      "step": 1701888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2126715597796854e-05,
+      "loss": 3.884,
+      "step": 1702400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2118329650286334e-05,
+      "loss": 3.8982,
+      "step": 1702912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2109960081579546e-05,
+      "loss": 3.8875,
+      "step": 1703424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2101574134069026e-05,
+      "loss": 3.8707,
+      "step": 1703936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2093188186558506e-05,
+      "loss": 3.9027,
+      "step": 1704448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2084802239047986e-05,
+      "loss": 3.8814,
+      "step": 1704960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2076416291537466e-05,
+      "loss": 3.8955,
+      "step": 1705472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2068030344026946e-05,
+      "loss": 3.8832,
+      "step": 1705984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2059660775320155e-05,
+      "loss": 3.8832,
+      "step": 1706496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.205127482780964e-05,
+      "loss": 3.8757,
+      "step": 1707008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.204290525910285e-05,
+      "loss": 3.8851,
+      "step": 1707520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2034519311592328e-05,
+      "loss": 3.8934,
+      "step": 1708032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2026133364081808e-05,
+      "loss": 3.8796,
+      "step": 1708544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2017747416571287e-05,
+      "loss": 3.895,
+      "step": 1709056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2009361469060767e-05,
+      "loss": 3.8882,
+      "step": 1709568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.2000975521550247e-05,
+      "loss": 3.8749,
+      "step": 1710080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.199258957403973e-05,
+      "loss": 3.902,
+      "step": 1710592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1984203626529207e-05,
+      "loss": 3.8715,
+      "step": 1711104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.197583405782242e-05,
+      "loss": 3.8681,
+      "step": 1711616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.19674481103119e-05,
+      "loss": 3.8886,
+      "step": 1712128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.195906216280138e-05,
+      "loss": 3.8979,
+      "step": 1712640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1950676215290856e-05,
+      "loss": 3.8753,
+      "step": 1713152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.194229026778034e-05,
+      "loss": 3.8893,
+      "step": 1713664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.193390432026982e-05,
+      "loss": 3.8696,
+      "step": 1714176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.19255183727593e-05,
+      "loss": 3.8826,
+      "step": 1714688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.191713242524878e-05,
+      "loss": 3.8902,
+      "step": 1715200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.190876285654199e-05,
+      "loss": 3.8856,
+      "step": 1715712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.190037690903147e-05,
+      "loss": 3.8889,
+      "step": 1716224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.189199096152095e-05,
+      "loss": 3.8889,
+      "step": 1716736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.188362139281416e-05,
+      "loss": 3.8977,
+      "step": 1717248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.187523544530364e-05,
+      "loss": 3.8759,
+      "step": 1717760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1866882255400583e-05,
+      "loss": 3.8872,
+      "step": 1718272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1858496307890063e-05,
+      "loss": 3.8883,
+      "step": 1718784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1850110360379546e-05,
+      "loss": 3.8785,
+      "step": 1719296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1841724412869026e-05,
+      "loss": 3.8879,
+      "step": 1719808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1833338465358506e-05,
+      "loss": 3.8867,
+      "step": 1720320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1824952517847982e-05,
+      "loss": 3.8972,
+      "step": 1720832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1816566570337462e-05,
+      "loss": 3.8921,
+      "step": 1721344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1808180622826942e-05,
+      "loss": 3.8726,
+      "step": 1721856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1799794675316422e-05,
+      "loss": 3.8755,
+      "step": 1722368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1791408727805902e-05,
+      "loss": 3.8896,
+      "step": 1722880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1783022780295385e-05,
+      "loss": 3.894,
+      "step": 1723392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1774636832784862e-05,
+      "loss": 3.8855,
+      "step": 1723904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.1766267264078075e-05,
+      "loss": 3.8809,
+      "step": 1724416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1757881316567554e-05,
+      "loss": 3.8889,
+      "step": 1724928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1749495369057034e-05,
+      "loss": 3.8851,
+      "step": 1725440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1741109421546514e-05,
+      "loss": 3.8847,
+      "step": 1725952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1732739852839727e-05,
+      "loss": 3.8779,
+      "step": 1726464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1724353905329207e-05,
+      "loss": 3.8948,
+      "step": 1726976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1715967957818687e-05,
+      "loss": 3.8898,
+      "step": 1727488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1707582010308163e-05,
+      "loss": 3.8917,
+      "step": 1728000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1699196062797643e-05,
+      "loss": 3.8784,
+      "step": 1728512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1690810115287123e-05,
+      "loss": 3.8856,
+      "step": 1729024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1682440546580336e-05,
+      "loss": 3.8832,
+      "step": 1729536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1674054599069816e-05,
+      "loss": 3.8917,
+      "step": 1730048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1665685030363028e-05,
+      "loss": 3.887,
+      "step": 1730560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1657299082852508e-05,
+      "loss": 3.8816,
+      "step": 1731072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1648913135341985e-05,
+      "loss": 3.8858,
+      "step": 1731584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1640527187831465e-05,
+      "loss": 3.8756,
+      "step": 1732096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1632141240320948e-05,
+      "loss": 3.878,
+      "step": 1732608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1623755292810428e-05,
+      "loss": 3.8746,
+      "step": 1733120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1615369345299908e-05,
+      "loss": 3.8768,
+      "step": 1733632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1606983397789388e-05,
+      "loss": 3.8822,
+      "step": 1734144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1598597450278864e-05,
+      "loss": 3.8926,
+      "step": 1734656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1590227881572077e-05,
+      "loss": 3.8752,
+      "step": 1735168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1581841934061557e-05,
+      "loss": 3.8806,
+      "step": 1735680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.157345598655104e-05,
+      "loss": 3.8819,
+      "step": 1736192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1565070039040517e-05,
+      "loss": 3.8887,
+      "step": 1736704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1556684091529997e-05,
+      "loss": 3.8801,
+      "step": 1737216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1548298144019477e-05,
+      "loss": 3.8918,
+      "step": 1737728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1539912196508957e-05,
+      "loss": 3.8871,
+      "step": 1738240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.153154262780217e-05,
+      "loss": 3.8869,
+      "step": 1738752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.152317305909538e-05,
+      "loss": 3.8859,
+      "step": 1739264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.151478711158486e-05,
+      "loss": 3.8852,
+      "step": 1739776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1506401164074338e-05,
+      "loss": 3.8843,
+      "step": 1740288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1498015216563818e-05,
+      "loss": 3.8848,
+      "step": 1740800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1489629269053298e-05,
+      "loss": 3.882,
+      "step": 1741312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1481243321542778e-05,
+      "loss": 3.8774,
+      "step": 1741824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.147285737403226e-05,
+      "loss": 3.8906,
+      "step": 1742336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.146447142652174e-05,
+      "loss": 3.8867,
+      "step": 1742848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.145608547901122e-05,
+      "loss": 3.8825,
+      "step": 1743360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1447699531500698e-05,
+      "loss": 3.8684,
+      "step": 1743872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.143932996279391e-05,
+      "loss": 3.8774,
+      "step": 1744384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.143094401528339e-05,
+      "loss": 3.8758,
+      "step": 1744896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.142255806777287e-05,
+      "loss": 3.889,
+      "step": 1745408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.141417212026235e-05,
+      "loss": 3.8929,
+      "step": 1745920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.140578617275183e-05,
+      "loss": 3.8945,
+      "step": 1746432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.139740022524131e-05,
+      "loss": 3.8717,
+      "step": 1746944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.138901427773079e-05,
+      "loss": 3.8935,
+      "step": 1747456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.138062833022027e-05,
+      "loss": 3.8863,
+      "step": 1747968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.137225876151348e-05,
+      "loss": 3.8858,
+      "step": 1748480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1363889192806695e-05,
+      "loss": 3.8786,
+      "step": 1748992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.135550324529617e-05,
+      "loss": 3.8894,
+      "step": 1749504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.134711729778565e-05,
+      "loss": 3.8773,
+      "step": 1750016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.133873135027513e-05,
+      "loss": 3.893,
+      "step": 1750528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.133034540276461e-05,
+      "loss": 3.8732,
+      "step": 1751040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1321975834057824e-05,
+      "loss": 3.8888,
+      "step": 1751552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1313589886547304e-05,
+      "loss": 3.8841,
+      "step": 1752064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1305203939036784e-05,
+      "loss": 3.8821,
+      "step": 1752576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1296817991526264e-05,
+      "loss": 3.8864,
+      "step": 1753088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1288432044015744e-05,
+      "loss": 3.8857,
+      "step": 1753600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1280046096505224e-05,
+      "loss": 3.8862,
+      "step": 1754112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.12716601489947e-05,
+      "loss": 3.8796,
+      "step": 1754624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.1263290580287916e-05,
+      "loss": 3.8935,
+      "step": 1755136
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.007114887237549,
+      "eval_runtime": 293.3044,
+      "eval_samples_per_second": 1301.007,
+      "eval_steps_per_second": 40.657,
+      "step": 1755360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1254904632777396e-05,
+      "loss": 3.8826,
+      "step": 1755648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1246518685266876e-05,
+      "loss": 3.8734,
+      "step": 1756160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1238132737756353e-05,
+      "loss": 3.8892,
+      "step": 1756672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1229746790245833e-05,
+      "loss": 3.8892,
+      "step": 1757184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1221360842735313e-05,
+      "loss": 3.8925,
+      "step": 1757696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1212974895224792e-05,
+      "loss": 3.881,
+      "step": 1758208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1204588947714272e-05,
+      "loss": 3.8858,
+      "step": 1758720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1196203000203756e-05,
+      "loss": 3.8717,
+      "step": 1759232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1187817052693236e-05,
+      "loss": 3.8847,
+      "step": 1759744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1179431105182712e-05,
+      "loss": 3.8839,
+      "step": 1760256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1171045157672192e-05,
+      "loss": 3.8854,
+      "step": 1760768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1162659210161672e-05,
+      "loss": 3.8889,
+      "step": 1761280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.115428964145488e-05,
+      "loss": 3.8837,
+      "step": 1761792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1145903693944365e-05,
+      "loss": 3.8811,
+      "step": 1762304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1137517746433845e-05,
+      "loss": 3.8727,
+      "step": 1762816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1129131798923325e-05,
+      "loss": 3.8697,
+      "step": 1763328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1120745851412805e-05,
+      "loss": 3.8829,
+      "step": 1763840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1112359903902284e-05,
+      "loss": 3.887,
+      "step": 1764352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1103973956391764e-05,
+      "loss": 3.8809,
+      "step": 1764864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.109558800888124e-05,
+      "loss": 3.8942,
+      "step": 1765376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1087202061370724e-05,
+      "loss": 3.8961,
+      "step": 1765888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1078816113860204e-05,
+      "loss": 3.8965,
+      "step": 1766400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1070430166349684e-05,
+      "loss": 3.8831,
+      "step": 1766912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1062044218839164e-05,
+      "loss": 3.8846,
+      "step": 1767424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1053691028936106e-05,
+      "loss": 3.8831,
+      "step": 1767936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1045305081425586e-05,
+      "loss": 3.8936,
+      "step": 1768448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1036919133915066e-05,
+      "loss": 3.8843,
+      "step": 1768960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1028533186404546e-05,
+      "loss": 3.8809,
+      "step": 1769472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1020147238894026e-05,
+      "loss": 3.8843,
+      "step": 1769984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1011761291383506e-05,
+      "loss": 3.8813,
+      "step": 1770496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1003391722676715e-05,
+      "loss": 3.8782,
+      "step": 1771008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0995005775166195e-05,
+      "loss": 3.8883,
+      "step": 1771520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0986619827655678e-05,
+      "loss": 3.8895,
+      "step": 1772032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0978233880145158e-05,
+      "loss": 3.8915,
+      "step": 1772544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0969864311438367e-05,
+      "loss": 3.8779,
+      "step": 1773056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0961478363927847e-05,
+      "loss": 3.8914,
+      "step": 1773568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0953092416417327e-05,
+      "loss": 3.8854,
+      "step": 1774080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0944706468906807e-05,
+      "loss": 3.8693,
+      "step": 1774592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.093633690020002e-05,
+      "loss": 3.8724,
+      "step": 1775104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.09279509526895e-05,
+      "loss": 3.8761,
+      "step": 1775616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.091956500517898e-05,
+      "loss": 3.882,
+      "step": 1776128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.091117905766846e-05,
+      "loss": 3.8905,
+      "step": 1776640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.090279311015794e-05,
+      "loss": 3.8852,
+      "step": 1777152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.089440716264742e-05,
+      "loss": 3.8752,
+      "step": 1777664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0886021215136896e-05,
+      "loss": 3.8957,
+      "step": 1778176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.087763526762638e-05,
+      "loss": 3.8794,
+      "step": 1778688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.086926569891959e-05,
+      "loss": 3.8951,
+      "step": 1779200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.08608961302128e-05,
+      "loss": 3.8861,
+      "step": 1779712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.085251018270228e-05,
+      "loss": 3.8604,
+      "step": 1780224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.084412423519176e-05,
+      "loss": 3.9012,
+      "step": 1780736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.083573828768124e-05,
+      "loss": 3.8746,
+      "step": 1781248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.082735234017072e-05,
+      "loss": 3.8863,
+      "step": 1781760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.08189663926602e-05,
+      "loss": 3.8784,
+      "step": 1782272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.081058044514968e-05,
+      "loss": 3.8741,
+      "step": 1782784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.080219449763916e-05,
+      "loss": 3.8771,
+      "step": 1783296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.079380855012864e-05,
+      "loss": 3.8753,
+      "step": 1783808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.078542260261812e-05,
+      "loss": 3.8906,
+      "step": 1784320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0777053033911333e-05,
+      "loss": 3.8709,
+      "step": 1784832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0768667086400813e-05,
+      "loss": 3.8884,
+      "step": 1785344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0760281138890293e-05,
+      "loss": 3.8843,
+      "step": 1785856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0751895191379773e-05,
+      "loss": 3.8663,
+      "step": 1786368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.074350924386925e-05,
+      "loss": 3.8904,
+      "step": 1786880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.073512329635873e-05,
+      "loss": 3.8748,
+      "step": 1787392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0726770106455674e-05,
+      "loss": 3.8643,
+      "step": 1787904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0718384158945154e-05,
+      "loss": 3.8781,
+      "step": 1788416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0709998211434634e-05,
+      "loss": 3.8891,
+      "step": 1788928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0701612263924114e-05,
+      "loss": 3.8714,
+      "step": 1789440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0693226316413594e-05,
+      "loss": 3.8853,
+      "step": 1789952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0684840368903074e-05,
+      "loss": 3.867,
+      "step": 1790464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.067645442139255e-05,
+      "loss": 3.8756,
+      "step": 1790976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0668068473882034e-05,
+      "loss": 3.8858,
+      "step": 1791488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0659682526371514e-05,
+      "loss": 3.8799,
+      "step": 1792000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0651312957664723e-05,
+      "loss": 3.8875,
+      "step": 1792512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0642927010154203e-05,
+      "loss": 3.8804,
+      "step": 1793024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0634541062643683e-05,
+      "loss": 3.8902,
+      "step": 1793536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0626155115133163e-05,
+      "loss": 3.8755,
+      "step": 1794048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0617785546426375e-05,
+      "loss": 3.8824,
+      "step": 1794560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0609399598915855e-05,
+      "loss": 3.8772,
+      "step": 1795072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0601013651405335e-05,
+      "loss": 3.8751,
+      "step": 1795584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0592627703894815e-05,
+      "loss": 3.8812,
+      "step": 1796096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0584241756384295e-05,
+      "loss": 3.8823,
+      "step": 1796608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0575855808873775e-05,
+      "loss": 3.8866,
+      "step": 1797120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0567469861363255e-05,
+      "loss": 3.8905,
+      "step": 1797632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0559083913852735e-05,
+      "loss": 3.8638,
+      "step": 1798144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0550714345145947e-05,
+      "loss": 3.8694,
+      "step": 1798656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0542328397635427e-05,
+      "loss": 3.8817,
+      "step": 1799168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0533942450124904e-05,
+      "loss": 3.8889,
+      "step": 1799680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0525556502614384e-05,
+      "loss": 3.8795,
+      "step": 1800192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0517186933907596e-05,
+      "loss": 3.874,
+      "step": 1800704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.050880098639708e-05,
+      "loss": 3.8881,
+      "step": 1801216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0500415038886556e-05,
+      "loss": 3.8761,
+      "step": 1801728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0492029091376036e-05,
+      "loss": 3.8797,
+      "step": 1802240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0483643143865516e-05,
+      "loss": 3.8724,
+      "step": 1802752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.047527357515873e-05,
+      "loss": 3.8877,
+      "step": 1803264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0466887627648205e-05,
+      "loss": 3.8876,
+      "step": 1803776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.045850168013769e-05,
+      "loss": 3.8867,
+      "step": 1804288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.045011573262717e-05,
+      "loss": 3.872,
+      "step": 1804800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.044172978511665e-05,
+      "loss": 3.8743,
+      "step": 1805312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.043334383760613e-05,
+      "loss": 3.8745,
+      "step": 1805824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.042495789009561e-05,
+      "loss": 3.8875,
+      "step": 1806336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0416571942585085e-05,
+      "loss": 3.8809,
+      "step": 1806848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.040821875268203e-05,
+      "loss": 3.8752,
+      "step": 1807360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.039983280517151e-05,
+      "loss": 3.8795,
+      "step": 1807872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.039144685766099e-05,
+      "loss": 3.8724,
+      "step": 1808384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.038306091015047e-05,
+      "loss": 3.8724,
+      "step": 1808896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.037467496263995e-05,
+      "loss": 3.8701,
+      "step": 1809408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.036630539393316e-05,
+      "loss": 3.8746,
+      "step": 1809920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0357919446422642e-05,
+      "loss": 3.878,
+      "step": 1810432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0349533498912122e-05,
+      "loss": 3.8833,
+      "step": 1810944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0341147551401602e-05,
+      "loss": 3.8741,
+      "step": 1811456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0332761603891082e-05,
+      "loss": 3.8669,
+      "step": 1811968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.032437565638056e-05,
+      "loss": 3.8777,
+      "step": 1812480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.031598970887004e-05,
+      "loss": 3.8826,
+      "step": 1812992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0307636518966984e-05,
+      "loss": 3.8772,
+      "step": 1813504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0299250571456464e-05,
+      "loss": 3.8819,
+      "step": 1814016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0290864623945944e-05,
+      "loss": 3.8789,
+      "step": 1814528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0282478676435424e-05,
+      "loss": 3.886,
+      "step": 1815040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0274092728924904e-05,
+      "loss": 3.877,
+      "step": 1815552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.026570678141438e-05,
+      "loss": 3.8775,
+      "step": 1816064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.025732083390386e-05,
+      "loss": 3.875,
+      "step": 1816576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0248934886393343e-05,
+      "loss": 3.8881,
+      "step": 1817088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0240548938882823e-05,
+      "loss": 3.8755,
+      "step": 1817600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0232162991372303e-05,
+      "loss": 3.8739,
+      "step": 1818112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0223777043861783e-05,
+      "loss": 3.8816,
+      "step": 1818624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0215391096351263e-05,
+      "loss": 3.8856,
+      "step": 1819136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0207021527644472e-05,
+      "loss": 3.8726,
+      "step": 1819648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0198635580133952e-05,
+      "loss": 3.8656,
+      "step": 1820160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0190266011427165e-05,
+      "loss": 3.871,
+      "step": 1820672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0181896442720377e-05,
+      "loss": 3.8682,
+      "step": 1821184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0173510495209854e-05,
+      "loss": 3.8812,
+      "step": 1821696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0165124547699334e-05,
+      "loss": 3.8903,
+      "step": 1822208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0156738600188814e-05,
+      "loss": 3.8892,
+      "step": 1822720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0148352652678297e-05,
+      "loss": 3.8627,
+      "step": 1823232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0139966705167777e-05,
+      "loss": 3.8896,
+      "step": 1823744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0131580757657257e-05,
+      "loss": 3.8811,
+      "step": 1824256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0123194810146737e-05,
+      "loss": 3.879,
+      "step": 1824768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0114808862636213e-05,
+      "loss": 3.8719,
+      "step": 1825280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0106422915125693e-05,
+      "loss": 3.8845,
+      "step": 1825792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0098036967615173e-05,
+      "loss": 3.8731,
+      "step": 1826304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0089651020104657e-05,
+      "loss": 3.8905,
+      "step": 1826816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0081265072594137e-05,
+      "loss": 3.8688,
+      "step": 1827328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0072879125083617e-05,
+      "loss": 3.8786,
+      "step": 1827840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0064493177573097e-05,
+      "loss": 3.8793,
+      "step": 1828352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0056123608866306e-05,
+      "loss": 3.88,
+      "step": 1828864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0047754040159518e-05,
+      "loss": 3.8775,
+      "step": 1829376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0039368092648998e-05,
+      "loss": 3.8786,
+      "step": 1829888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0030982145138478e-05,
+      "loss": 3.8834,
+      "step": 1830400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0022596197627958e-05,
+      "loss": 3.8773,
+      "step": 1830912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0014226628921167e-05,
+      "loss": 3.8858,
+      "step": 1831424
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0050835609436035,
+      "eval_runtime": 290.3475,
+      "eval_samples_per_second": 1314.256,
+      "eval_steps_per_second": 41.071,
+      "step": 1831680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0005840681410647e-05,
+      "loss": 3.8734,
+      "step": 1831936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9997454733900127e-05,
+      "loss": 3.8703,
+      "step": 1832448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9989068786389607e-05,
+      "loss": 3.8789,
+      "step": 1832960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.998068283887909e-05,
+      "loss": 3.8813,
+      "step": 1833472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.99723132701723e-05,
+      "loss": 3.8906,
+      "step": 1833984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.996392732266178e-05,
+      "loss": 3.8746,
+      "step": 1834496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.995554137515126e-05,
+      "loss": 3.8832,
+      "step": 1835008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.994715542764074e-05,
+      "loss": 3.8697,
+      "step": 1835520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.993876948013022e-05,
+      "loss": 3.8746,
+      "step": 1836032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9930399911423432e-05,
+      "loss": 3.8771,
+      "step": 1836544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9922013963912912e-05,
+      "loss": 3.88,
+      "step": 1837056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.991362801640239e-05,
+      "loss": 3.8818,
+      "step": 1837568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.99052584476956e-05,
+      "loss": 3.881,
+      "step": 1838080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.989687250018508e-05,
+      "loss": 3.872,
+      "step": 1838592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.988848655267456e-05,
+      "loss": 3.8718,
+      "step": 1839104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9880100605164044e-05,
+      "loss": 3.8637,
+      "step": 1839616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.987171465765352e-05,
+      "loss": 3.8789,
+      "step": 1840128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9863328710143e-05,
+      "loss": 3.8849,
+      "step": 1840640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.985494276263248e-05,
+      "loss": 3.8717,
+      "step": 1841152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.984655681512196e-05,
+      "loss": 3.8894,
+      "step": 1841664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9838187246415173e-05,
+      "loss": 3.8893,
+      "step": 1842176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9829801298904653e-05,
+      "loss": 3.8896,
+      "step": 1842688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9821431730197865e-05,
+      "loss": 3.8855,
+      "step": 1843200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9813045782687342e-05,
+      "loss": 3.8754,
+      "step": 1843712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9804659835176822e-05,
+      "loss": 3.8775,
+      "step": 1844224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9796273887666302e-05,
+      "loss": 3.8836,
+      "step": 1844736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9787887940155782e-05,
+      "loss": 3.8837,
+      "step": 1845248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9779501992645265e-05,
+      "loss": 3.876,
+      "step": 1845760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9771116045134745e-05,
+      "loss": 3.8785,
+      "step": 1846272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9762730097624222e-05,
+      "loss": 3.8767,
+      "step": 1846784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9754360528917434e-05,
+      "loss": 3.8715,
+      "step": 1847296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9745974581406914e-05,
+      "loss": 3.8838,
+      "step": 1847808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9737588633896394e-05,
+      "loss": 3.8815,
+      "step": 1848320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9729202686385874e-05,
+      "loss": 3.8868,
+      "step": 1848832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9720833117679087e-05,
+      "loss": 3.8753,
+      "step": 1849344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9712447170168566e-05,
+      "loss": 3.8833,
+      "step": 1849856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9704061222658046e-05,
+      "loss": 3.8798,
+      "step": 1850368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9695675275147523e-05,
+      "loss": 3.8626,
+      "step": 1850880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9687289327637003e-05,
+      "loss": 3.8718,
+      "step": 1851392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9678919758930215e-05,
+      "loss": 3.8661,
+      "step": 1851904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9670533811419695e-05,
+      "loss": 3.8736,
+      "step": 1852416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9662147863909175e-05,
+      "loss": 3.8896,
+      "step": 1852928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9653778295202388e-05,
+      "loss": 3.8774,
+      "step": 1853440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9645392347691868e-05,
+      "loss": 3.871,
+      "step": 1853952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9637006400181344e-05,
+      "loss": 3.8838,
+      "step": 1854464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9628620452670828e-05,
+      "loss": 3.8742,
+      "step": 1854976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9620234505160308e-05,
+      "loss": 3.8884,
+      "step": 1855488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9611848557649788e-05,
+      "loss": 3.8783,
+      "step": 1856000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9603462610139268e-05,
+      "loss": 3.8537,
+      "step": 1856512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9595076662628748e-05,
+      "loss": 3.8948,
+      "step": 1857024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9586690715118227e-05,
+      "loss": 3.8711,
+      "step": 1857536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9578304767607704e-05,
+      "loss": 3.8814,
+      "step": 1858048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9569918820097184e-05,
+      "loss": 3.8724,
+      "step": 1858560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9561532872586667e-05,
+      "loss": 3.8685,
+      "step": 1859072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9553163303879876e-05,
+      "loss": 3.87,
+      "step": 1859584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9544777356369356e-05,
+      "loss": 3.8718,
+      "step": 1860096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9536391408858836e-05,
+      "loss": 3.8889,
+      "step": 1860608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9528005461348316e-05,
+      "loss": 3.8638,
+      "step": 1861120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951963589264153e-05,
+      "loss": 3.8872,
+      "step": 1861632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951126632393474e-05,
+      "loss": 3.878,
+      "step": 1862144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.950288037642422e-05,
+      "loss": 3.8579,
+      "step": 1862656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.94944944289137e-05,
+      "loss": 3.8885,
+      "step": 1863168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9486108481403178e-05,
+      "loss": 3.8693,
+      "step": 1863680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9477722533892658e-05,
+      "loss": 3.8575,
+      "step": 1864192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9469336586382138e-05,
+      "loss": 3.8719,
+      "step": 1864704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.946095063887162e-05,
+      "loss": 3.8837,
+      "step": 1865216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.945258107016483e-05,
+      "loss": 3.8691,
+      "step": 1865728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.944419512265431e-05,
+      "loss": 3.8777,
+      "step": 1866240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.943580917514379e-05,
+      "loss": 3.8655,
+      "step": 1866752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.942742322763327e-05,
+      "loss": 3.8664,
+      "step": 1867264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.941903728012275e-05,
+      "loss": 3.8783,
+      "step": 1867776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.941065133261223e-05,
+      "loss": 3.876,
+      "step": 1868288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9402281763905442e-05,
+      "loss": 3.8787,
+      "step": 1868800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9393895816394922e-05,
+      "loss": 3.8755,
+      "step": 1869312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9385509868884402e-05,
+      "loss": 3.8866,
+      "step": 1869824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9377123921373882e-05,
+      "loss": 3.8707,
+      "step": 1870336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.936875435266709e-05,
+      "loss": 3.8786,
+      "step": 1870848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9360368405156575e-05,
+      "loss": 3.8703,
+      "step": 1871360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9351982457646055e-05,
+      "loss": 3.8711,
+      "step": 1871872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.934359651013553e-05,
+      "loss": 3.877,
+      "step": 1872384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9335226941428744e-05,
+      "loss": 3.8731,
+      "step": 1872896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9326840993918224e-05,
+      "loss": 3.8819,
+      "step": 1873408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9318455046407704e-05,
+      "loss": 3.8886,
+      "step": 1873920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9310069098897184e-05,
+      "loss": 3.8579,
+      "step": 1874432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9301683151386664e-05,
+      "loss": 3.8644,
+      "step": 1874944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9293313582679876e-05,
+      "loss": 3.8779,
+      "step": 1875456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9284927635169356e-05,
+      "loss": 3.8793,
+      "step": 1875968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9276541687658833e-05,
+      "loss": 3.8774,
+      "step": 1876480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9268155740148312e-05,
+      "loss": 3.8718,
+      "step": 1876992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.925978617144153e-05,
+      "loss": 3.8787,
+      "step": 1877504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9251400223931005e-05,
+      "loss": 3.8733,
+      "step": 1878016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9243014276420485e-05,
+      "loss": 3.8719,
+      "step": 1878528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9234644707713697e-05,
+      "loss": 3.8719,
+      "step": 1879040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9226258760203177e-05,
+      "loss": 3.8813,
+      "step": 1879552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9217872812692654e-05,
+      "loss": 3.8832,
+      "step": 1880064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9209486865182137e-05,
+      "loss": 3.8821,
+      "step": 1880576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9201100917671617e-05,
+      "loss": 3.8666,
+      "step": 1881088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9192714970161097e-05,
+      "loss": 3.8727,
+      "step": 1881600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9184329022650577e-05,
+      "loss": 3.8696,
+      "step": 1882112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9175943075140057e-05,
+      "loss": 3.8791,
+      "step": 1882624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9167573506433266e-05,
+      "loss": 3.8779,
+      "step": 1883136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9159187558922746e-05,
+      "loss": 3.8665,
+      "step": 1883648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.915080161141223e-05,
+      "loss": 3.8747,
+      "step": 1884160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.914241566390171e-05,
+      "loss": 3.8629,
+      "step": 1884672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9134029716391186e-05,
+      "loss": 3.8694,
+      "step": 1885184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9125643768880666e-05,
+      "loss": 3.8629,
+      "step": 1885696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.911727420017388e-05,
+      "loss": 3.872,
+      "step": 1886208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.910888825266336e-05,
+      "loss": 3.8728,
+      "step": 1886720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.910050230515284e-05,
+      "loss": 3.8773,
+      "step": 1887232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9092116357642318e-05,
+      "loss": 3.8666,
+      "step": 1887744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9083730410131798e-05,
+      "loss": 3.8665,
+      "step": 1888256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9075344462621278e-05,
+      "loss": 3.8685,
+      "step": 1888768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9066958515110758e-05,
+      "loss": 3.8796,
+      "step": 1889280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9058572567600238e-05,
+      "loss": 3.8719,
+      "step": 1889792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9050202998893447e-05,
+      "loss": 3.8733,
+      "step": 1890304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.904181705138293e-05,
+      "loss": 3.8736,
+      "step": 1890816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.903343110387241e-05,
+      "loss": 3.8859,
+      "step": 1891328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.902504515636189e-05,
+      "loss": 3.8718,
+      "step": 1891840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9016659208851367e-05,
+      "loss": 3.8721,
+      "step": 1892352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9008273261340847e-05,
+      "loss": 3.8747,
+      "step": 1892864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8999887313830327e-05,
+      "loss": 3.8792,
+      "step": 1893376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8991501366319807e-05,
+      "loss": 3.8706,
+      "step": 1893888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.898311541880929e-05,
+      "loss": 3.8695,
+      "step": 1894400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8974762228906232e-05,
+      "loss": 3.8756,
+      "step": 1894912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8966376281395712e-05,
+      "loss": 3.8804,
+      "step": 1895424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.895799033388519e-05,
+      "loss": 3.8695,
+      "step": 1895936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.894960438637467e-05,
+      "loss": 3.8615,
+      "step": 1896448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8941234817667884e-05,
+      "loss": 3.8644,
+      "step": 1896960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8932865248961093e-05,
+      "loss": 3.8617,
+      "step": 1897472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8924479301450573e-05,
+      "loss": 3.8792,
+      "step": 1897984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8916093353940053e-05,
+      "loss": 3.8794,
+      "step": 1898496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8907707406429533e-05,
+      "loss": 3.8852,
+      "step": 1899008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8899321458919013e-05,
+      "loss": 3.8591,
+      "step": 1899520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8890935511408493e-05,
+      "loss": 3.8831,
+      "step": 1900032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8882549563897973e-05,
+      "loss": 3.8752,
+      "step": 1900544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8874163616387453e-05,
+      "loss": 3.8759,
+      "step": 1901056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8865794047680662e-05,
+      "loss": 3.8695,
+      "step": 1901568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8857408100170142e-05,
+      "loss": 3.8721,
+      "step": 1902080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8849022152659622e-05,
+      "loss": 3.8693,
+      "step": 1902592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8840636205149105e-05,
+      "loss": 3.885,
+      "step": 1903104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8832266636442314e-05,
+      "loss": 3.8621,
+      "step": 1903616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8823880688931794e-05,
+      "loss": 3.8743,
+      "step": 1904128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8815494741421274e-05,
+      "loss": 3.8772,
+      "step": 1904640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8807108793910754e-05,
+      "loss": 3.877,
+      "step": 1905152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8798722846400234e-05,
+      "loss": 3.8667,
+      "step": 1905664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8790336898889714e-05,
+      "loss": 3.874,
+      "step": 1906176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8781950951379194e-05,
+      "loss": 3.8786,
+      "step": 1906688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8773565003868674e-05,
+      "loss": 3.8706,
+      "step": 1907200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8765195435161887e-05,
+      "loss": 3.8802,
+      "step": 1907712
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.00334358215332,
+      "eval_runtime": 291.0073,
+      "eval_samples_per_second": 1311.276,
+      "eval_steps_per_second": 40.978,
+      "step": 1908000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8756809487651367e-05,
+      "loss": 3.8706,
+      "step": 1908224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8748423540140843e-05,
+      "loss": 3.8633,
+      "step": 1908736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8740037592630323e-05,
+      "loss": 3.8733,
+      "step": 1909248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8731651645119806e-05,
+      "loss": 3.8787,
+      "step": 1909760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.872328207641302e-05,
+      "loss": 3.8828,
+      "step": 1910272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8714896128902495e-05,
+      "loss": 3.8695,
+      "step": 1910784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8706510181391975e-05,
+      "loss": 3.8766,
+      "step": 1911296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8698124233881455e-05,
+      "loss": 3.861,
+      "step": 1911808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8689738286370935e-05,
+      "loss": 3.8711,
+      "step": 1912320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8681352338860415e-05,
+      "loss": 3.8731,
+      "step": 1912832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.86729663913499e-05,
+      "loss": 3.8756,
+      "step": 1913344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.866458044383938e-05,
+      "loss": 3.8779,
+      "step": 1913856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8656227253936317e-05,
+      "loss": 3.8807,
+      "step": 1914368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8647841306425797e-05,
+      "loss": 3.8658,
+      "step": 1914880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.863947173771901e-05,
+      "loss": 3.8644,
+      "step": 1915392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8631085790208493e-05,
+      "loss": 3.8553,
+      "step": 1915904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.862269984269797e-05,
+      "loss": 3.877,
+      "step": 1916416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.861431389518745e-05,
+      "loss": 3.8792,
+      "step": 1916928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.860594432648066e-05,
+      "loss": 3.8641,
+      "step": 1917440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8597558378970138e-05,
+      "loss": 3.8794,
+      "step": 1917952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.858917243145962e-05,
+      "loss": 3.8888,
+      "step": 1918464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.85807864839491e-05,
+      "loss": 3.8843,
+      "step": 1918976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.857240053643858e-05,
+      "loss": 3.8768,
+      "step": 1919488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.856401458892806e-05,
+      "loss": 3.8711,
+      "step": 1920000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.855562864141754e-05,
+      "loss": 3.8765,
+      "step": 1920512
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.854724269390702e-05,
+      "loss": 3.8742,
+      "step": 1921024
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8538856746396498e-05,
+      "loss": 3.8793,
+      "step": 1921536
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8530470798885978e-05,
+      "loss": 3.8762,
+      "step": 1922048
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.852208485137546e-05,
+      "loss": 3.8698,
+      "step": 1922560
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.851369890386494e-05,
+      "loss": 3.8674,
+      "step": 1923072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.850532933515815e-05,
+      "loss": 3.8697,
+      "step": 1923584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.849694338764763e-05,
+      "loss": 3.8781,
+      "step": 1924096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.848855744013711e-05,
+      "loss": 3.8812,
+      "step": 1924608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.848017149262659e-05,
+      "loss": 3.8807,
+      "step": 1925120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.847178554511607e-05,
+      "loss": 3.8697,
+      "step": 1925632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8463415976409283e-05,
+      "loss": 3.8725,
+      "step": 1926144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8455046407702495e-05,
+      "loss": 3.8783,
+      "step": 1926656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.844666046019197e-05,
+      "loss": 3.8606,
+      "step": 1927168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.843827451268145e-05,
+      "loss": 3.8654,
+      "step": 1927680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.842988856517093e-05,
+      "loss": 3.8652,
+      "step": 1928192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8421502617660415e-05,
+      "loss": 3.8657,
+      "step": 1928704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8413116670149895e-05,
+      "loss": 3.8825,
+      "step": 1929216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8404730722639375e-05,
+      "loss": 3.8716,
+      "step": 1929728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8396344775128855e-05,
+      "loss": 3.8688,
+      "step": 1930240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8387975206422064e-05,
+      "loss": 3.8789,
+      "step": 1930752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8379589258911544e-05,
+      "loss": 3.8668,
+      "step": 1931264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8371219690204756e-05,
+      "loss": 3.8787,
+      "step": 1931776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8362833742694236e-05,
+      "loss": 3.8832,
+      "step": 1932288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8354447795183716e-05,
+      "loss": 3.8487,
+      "step": 1932800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8346061847673196e-05,
+      "loss": 3.8844,
+      "step": 1933312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8337675900162676e-05,
+      "loss": 3.8677,
+      "step": 1933824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8329289952652153e-05,
+      "loss": 3.8745,
+      "step": 1934336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8320904005141633e-05,
+      "loss": 3.8679,
+      "step": 1934848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.831253443643485e-05,
+      "loss": 3.8629,
+      "step": 1935360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.830414848892433e-05,
+      "loss": 3.8658,
+      "step": 1935872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8295762541413805e-05,
+      "loss": 3.8648,
+      "step": 1936384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8287376593903285e-05,
+      "loss": 3.885,
+      "step": 1936896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8278990646392765e-05,
+      "loss": 3.857,
+      "step": 1937408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8270604698882245e-05,
+      "loss": 3.8817,
+      "step": 1937920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8262235130175457e-05,
+      "loss": 3.8722,
+      "step": 1938432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8253849182664937e-05,
+      "loss": 3.853,
+      "step": 1938944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8245463235154417e-05,
+      "loss": 3.8815,
+      "step": 1939456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8237077287643897e-05,
+      "loss": 3.8626,
+      "step": 1939968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8228691340133377e-05,
+      "loss": 3.8519,
+      "step": 1940480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8220321771426586e-05,
+      "loss": 3.8633,
+      "step": 1940992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.821193582391607e-05,
+      "loss": 3.8803,
+      "step": 1941504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.820354987640555e-05,
+      "loss": 3.8589,
+      "step": 1942016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.819516392889503e-05,
+      "loss": 3.8749,
+      "step": 1942528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.818679436018824e-05,
+      "loss": 3.8594,
+      "step": 1943040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.817840841267772e-05,
+      "loss": 3.8585,
+      "step": 1943552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.81700224651672e-05,
+      "loss": 3.8721,
+      "step": 1944064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.816163651765668e-05,
+      "loss": 3.8735,
+      "step": 1944576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.815325057014616e-05,
+      "loss": 3.8731,
+      "step": 1945088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.814486462263564e-05,
+      "loss": 3.8721,
+      "step": 1945600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.813647867512512e-05,
+      "loss": 3.8762,
+      "step": 1946112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.81280927276146e-05,
+      "loss": 3.8684,
+      "step": 1946624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8119723158907807e-05,
+      "loss": 3.8716,
+      "step": 1947136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8111337211397287e-05,
+      "loss": 3.8599,
+      "step": 1947648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.810295126388677e-05,
+      "loss": 3.8673,
+      "step": 1948160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.809456531637625e-05,
+      "loss": 3.8719,
+      "step": 1948672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.808617936886573e-05,
+      "loss": 3.8654,
+      "step": 1949184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.807780980015894e-05,
+      "loss": 3.8774,
+      "step": 1949696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.806942385264842e-05,
+      "loss": 3.8804,
+      "step": 1950208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.80610379051379e-05,
+      "loss": 3.8609,
+      "step": 1950720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.805265195762738e-05,
+      "loss": 3.8579,
+      "step": 1951232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8044266010116863e-05,
+      "loss": 3.8716,
+      "step": 1951744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8035896441410072e-05,
+      "loss": 3.8735,
+      "step": 1952256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8027510493899552e-05,
+      "loss": 3.8718,
+      "step": 1952768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8019124546389032e-05,
+      "loss": 3.8628,
+      "step": 1953280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8010738598878512e-05,
+      "loss": 3.8758,
+      "step": 1953792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8002352651367992e-05,
+      "loss": 3.8646,
+      "step": 1954304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7993983082661204e-05,
+      "loss": 3.8674,
+      "step": 1954816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7985597135150684e-05,
+      "loss": 3.866,
+      "step": 1955328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7977227566443893e-05,
+      "loss": 3.8751,
+      "step": 1955840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7968841618933373e-05,
+      "loss": 3.874,
+      "step": 1956352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7960455671422853e-05,
+      "loss": 3.8784,
+      "step": 1956864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7952069723912333e-05,
+      "loss": 3.8593,
+      "step": 1957376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7943683776401813e-05,
+      "loss": 3.8696,
+      "step": 1957888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7935297828891293e-05,
+      "loss": 3.8621,
+      "step": 1958400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7926928260184506e-05,
+      "loss": 3.8751,
+      "step": 1958912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7918542312673986e-05,
+      "loss": 3.8719,
+      "step": 1959424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7910156365163462e-05,
+      "loss": 3.8669,
+      "step": 1959936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7901770417652942e-05,
+      "loss": 3.87,
+      "step": 1960448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7893384470142425e-05,
+      "loss": 3.8584,
+      "step": 1960960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7884998522631905e-05,
+      "loss": 3.8633,
+      "step": 1961472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7876612575121385e-05,
+      "loss": 3.8585,
+      "step": 1961984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7868226627610865e-05,
+      "loss": 3.8674,
+      "step": 1962496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7859857058904074e-05,
+      "loss": 3.8682,
+      "step": 1963008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7851487490197287e-05,
+      "loss": 3.8709,
+      "step": 1963520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7843101542686767e-05,
+      "loss": 3.8607,
+      "step": 1964032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7834715595176247e-05,
+      "loss": 3.8596,
+      "step": 1964544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7826329647665727e-05,
+      "loss": 3.8631,
+      "step": 1965056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7817943700155207e-05,
+      "loss": 3.8733,
+      "step": 1965568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7809557752644687e-05,
+      "loss": 3.868,
+      "step": 1966080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7801171805134167e-05,
+      "loss": 3.8691,
+      "step": 1966592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.779280223642738e-05,
+      "loss": 3.8683,
+      "step": 1967104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.778441628891686e-05,
+      "loss": 3.8775,
+      "step": 1967616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.777603034140634e-05,
+      "loss": 3.8696,
+      "step": 1968128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7767644393895816e-05,
+      "loss": 3.8681,
+      "step": 1968640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7759258446385296e-05,
+      "loss": 3.8678,
+      "step": 1969152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7750872498874776e-05,
+      "loss": 3.8729,
+      "step": 1969664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7742486551364255e-05,
+      "loss": 3.8684,
+      "step": 1970176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.773410060385374e-05,
+      "loss": 3.8644,
+      "step": 1970688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.772571465634322e-05,
+      "loss": 3.8643,
+      "step": 1971200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.77173287088327e-05,
+      "loss": 3.8793,
+      "step": 1971712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7708959140125908e-05,
+      "loss": 3.8625,
+      "step": 1972224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7700589571419117e-05,
+      "loss": 3.8562,
+      "step": 1972736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.76922036239086e-05,
+      "loss": 3.8572,
+      "step": 1973248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.768381767639808e-05,
+      "loss": 3.8558,
+      "step": 1973760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.767543172888756e-05,
+      "loss": 3.8806,
+      "step": 1974272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.766704578137704e-05,
+      "loss": 3.8688,
+      "step": 1974784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.765865983386652e-05,
+      "loss": 3.8823,
+      "step": 1975296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7650273886355997e-05,
+      "loss": 3.8584,
+      "step": 1975808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7641887938845477e-05,
+      "loss": 3.8751,
+      "step": 1976320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.763351837013869e-05,
+      "loss": 3.8703,
+      "step": 1976832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7625132422628172e-05,
+      "loss": 3.8723,
+      "step": 1977344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.761674647511765e-05,
+      "loss": 3.8658,
+      "step": 1977856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.760836052760713e-05,
+      "loss": 3.8649,
+      "step": 1978368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.759997458009661e-05,
+      "loss": 3.8613,
+      "step": 1978880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.759158863258609e-05,
+      "loss": 3.8815,
+      "step": 1979392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.758320268507557e-05,
+      "loss": 3.8556,
+      "step": 1979904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.757481673756505e-05,
+      "loss": 3.8704,
+      "step": 1980416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.756644716885826e-05,
+      "loss": 3.8754,
+      "step": 1980928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.755806122134774e-05,
+      "loss": 3.8663,
+      "step": 1981440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.754967527383722e-05,
+      "loss": 3.862,
+      "step": 1981952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.75412893263267e-05,
+      "loss": 3.8699,
+      "step": 1982464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7532903378816178e-05,
+      "loss": 3.8722,
+      "step": 1982976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7524550188913123e-05,
+      "loss": 3.8702,
+      "step": 1983488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7516164241402603e-05,
+      "loss": 3.8737,
+      "step": 1984000
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.001668453216553,
+      "eval_runtime": 291.8071,
+      "eval_samples_per_second": 1307.683,
+      "eval_steps_per_second": 40.866,
+      "step": 1984320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7507778293892083e-05,
+      "loss": 3.8591,
+      "step": 1984512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7499392346381563e-05,
+      "loss": 3.8587,
+      "step": 1985024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7491006398871043e-05,
+      "loss": 3.8691,
+      "step": 1985536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7482620451360523e-05,
+      "loss": 3.8725,
+      "step": 1986048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7474234503850002e-05,
+      "loss": 3.881,
+      "step": 1986560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7465848556339482e-05,
+      "loss": 3.865,
+      "step": 1987072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7457462608828962e-05,
+      "loss": 3.8701,
+      "step": 1987584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7449076661318442e-05,
+      "loss": 3.8619,
+      "step": 1988096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7440690713807922e-05,
+      "loss": 3.8635,
+      "step": 1988608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7432304766297402e-05,
+      "loss": 3.8664,
+      "step": 1989120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7423918818786882e-05,
+      "loss": 3.8655,
+      "step": 1989632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7415532871276362e-05,
+      "loss": 3.8756,
+      "step": 1990144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7407163302569575e-05,
+      "loss": 3.8758,
+      "step": 1990656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7398777355059055e-05,
+      "loss": 3.8594,
+      "step": 1991168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7390391407548535e-05,
+      "loss": 3.8612,
+      "step": 1991680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.738200546003801e-05,
+      "loss": 3.8499,
+      "step": 1992192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.737361951252749e-05,
+      "loss": 3.8743,
+      "step": 1992704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.736523356501697e-05,
+      "loss": 3.8698,
+      "step": 1993216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.735684761750645e-05,
+      "loss": 3.8615,
+      "step": 1993728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7348461669995934e-05,
+      "loss": 3.8709,
+      "step": 1994240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7340075722485414e-05,
+      "loss": 3.8871,
+      "step": 1994752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7331722532582356e-05,
+      "loss": 3.8777,
+      "step": 1995264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7323336585071832e-05,
+      "loss": 3.8727,
+      "step": 1995776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7314950637561312e-05,
+      "loss": 3.8655,
+      "step": 1996288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7306564690050796e-05,
+      "loss": 3.8752,
+      "step": 1996800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7298178742540276e-05,
+      "loss": 3.8682,
+      "step": 1997312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7289792795029756e-05,
+      "loss": 3.8705,
+      "step": 1997824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7281406847519236e-05,
+      "loss": 3.8707,
+      "step": 1998336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7273020900008716e-05,
+      "loss": 3.8666,
+      "step": 1998848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7264651331301925e-05,
+      "loss": 3.8641,
+      "step": 1999360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7256265383791405e-05,
+      "loss": 3.8628,
+      "step": 1999872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7247879436280888e-05,
+      "loss": 3.8702,
+      "step": 2000384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.723952624637783e-05,
+      "loss": 3.8706,
+      "step": 2000896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7231140298867306e-05,
+      "loss": 3.8772,
+      "step": 2001408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7222754351356786e-05,
+      "loss": 3.8679,
+      "step": 2001920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7214368403846266e-05,
+      "loss": 3.8668,
+      "step": 2002432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.720598245633575e-05,
+      "loss": 3.8734,
+      "step": 2002944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.719759650882523e-05,
+      "loss": 3.8579,
+      "step": 2003456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.718922694011844e-05,
+      "loss": 3.8614,
+      "step": 2003968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.718084099260792e-05,
+      "loss": 3.8595,
+      "step": 2004480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.71724550450974e-05,
+      "loss": 3.8607,
+      "step": 2004992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.716406909758688e-05,
+      "loss": 3.8744,
+      "step": 2005504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.715568315007636e-05,
+      "loss": 3.8688,
+      "step": 2006016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7147297202565838e-05,
+      "loss": 3.8686,
+      "step": 2006528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7138911255055318e-05,
+      "loss": 3.8674,
+      "step": 2007040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7130525307544798e-05,
+      "loss": 3.8651,
+      "step": 2007552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7122139360034278e-05,
+      "loss": 3.8766,
+      "step": 2008064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7113753412523758e-05,
+      "loss": 3.8753,
+      "step": 2008576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7105367465013238e-05,
+      "loss": 3.8435,
+      "step": 2009088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7096981517502718e-05,
+      "loss": 3.8795,
+      "step": 2009600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7088595569992198e-05,
+      "loss": 3.8611,
+      "step": 2010112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.708022600128541e-05,
+      "loss": 3.8735,
+      "step": 2010624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.707184005377489e-05,
+      "loss": 3.8661,
+      "step": 2011136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.706345410626437e-05,
+      "loss": 3.8547,
+      "step": 2011648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7055068158753847e-05,
+      "loss": 3.8644,
+      "step": 2012160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.704669859004706e-05,
+      "loss": 3.8576,
+      "step": 2012672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7038312642536543e-05,
+      "loss": 3.8758,
+      "step": 2013184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7029943073829752e-05,
+      "loss": 3.8568,
+      "step": 2013696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7021557126319232e-05,
+      "loss": 3.8734,
+      "step": 2014208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7013171178808712e-05,
+      "loss": 3.8683,
+      "step": 2014720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7004785231298192e-05,
+      "loss": 3.8536,
+      "step": 2015232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.699639928378767e-05,
+      "loss": 3.8735,
+      "step": 2015744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6988029715080884e-05,
+      "loss": 3.8567,
+      "step": 2016256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6979643767570364e-05,
+      "loss": 3.8511,
+      "step": 2016768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6971257820059844e-05,
+      "loss": 3.8553,
+      "step": 2017280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.696287187254932e-05,
+      "loss": 3.8758,
+      "step": 2017792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6954502303842533e-05,
+      "loss": 3.8559,
+      "step": 2018304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6946116356332013e-05,
+      "loss": 3.8732,
+      "step": 2018816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6937730408821493e-05,
+      "loss": 3.8493,
+      "step": 2019328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6929344461310973e-05,
+      "loss": 3.8595,
+      "step": 2019840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6920958513800453e-05,
+      "loss": 3.8661,
+      "step": 2020352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6912572566289933e-05,
+      "loss": 3.8664,
+      "step": 2020864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6904186618779413e-05,
+      "loss": 3.8692,
+      "step": 2021376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6895800671268893e-05,
+      "loss": 3.8714,
+      "step": 2021888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6887414723758373e-05,
+      "loss": 3.8713,
+      "step": 2022400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6879028776247853e-05,
+      "loss": 3.8606,
+      "step": 2022912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6870642828737333e-05,
+      "loss": 3.8668,
+      "step": 2023424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6862256881226813e-05,
+      "loss": 3.8581,
+      "step": 2023936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6853887312520022e-05,
+      "loss": 3.863,
+      "step": 2024448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.68455013650095e-05,
+      "loss": 3.8632,
+      "step": 2024960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.683711541749898e-05,
+      "loss": 3.8587,
+      "step": 2025472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6828745848792198e-05,
+      "loss": 3.8714,
+      "step": 2025984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6820359901281674e-05,
+      "loss": 3.8758,
+      "step": 2026496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6811973953771154e-05,
+      "loss": 3.8616,
+      "step": 2027008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6803588006260634e-05,
+      "loss": 3.8479,
+      "step": 2027520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6795218437553846e-05,
+      "loss": 3.868,
+      "step": 2028032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6786832490043326e-05,
+      "loss": 3.8675,
+      "step": 2028544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6778446542532806e-05,
+      "loss": 3.8706,
+      "step": 2029056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6770060595022286e-05,
+      "loss": 3.8545,
+      "step": 2029568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6761674647511766e-05,
+      "loss": 3.8728,
+      "step": 2030080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6753288700001246e-05,
+      "loss": 3.8623,
+      "step": 2030592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6744902752490726e-05,
+      "loss": 3.8634,
+      "step": 2031104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6736516804980203e-05,
+      "loss": 3.8593,
+      "step": 2031616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6728130857469683e-05,
+      "loss": 3.8713,
+      "step": 2032128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6719744909959166e-05,
+      "loss": 3.8701,
+      "step": 2032640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.671137534125238e-05,
+      "loss": 3.8731,
+      "step": 2033152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6703005772545588e-05,
+      "loss": 3.8547,
+      "step": 2033664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6694619825035068e-05,
+      "loss": 3.8649,
+      "step": 2034176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6686233877524548e-05,
+      "loss": 3.859,
+      "step": 2034688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6677847930014028e-05,
+      "loss": 3.8674,
+      "step": 2035200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6669461982503507e-05,
+      "loss": 3.8672,
+      "step": 2035712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6661076034992987e-05,
+      "loss": 3.8626,
+      "step": 2036224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6652690087482467e-05,
+      "loss": 3.8627,
+      "step": 2036736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6644304139971947e-05,
+      "loss": 3.8528,
+      "step": 2037248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6635934571265156e-05,
+      "loss": 3.8628,
+      "step": 2037760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6627548623754636e-05,
+      "loss": 3.8497,
+      "step": 2038272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6619179055047852e-05,
+      "loss": 3.8615,
+      "step": 2038784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.661079310753733e-05,
+      "loss": 3.8591,
+      "step": 2039296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.660240716002681e-05,
+      "loss": 3.8697,
+      "step": 2039808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.659402121251629e-05,
+      "loss": 3.8552,
+      "step": 2040320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.658563526500577e-05,
+      "loss": 3.856,
+      "step": 2040832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.657724931749525e-05,
+      "loss": 3.8559,
+      "step": 2041344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.656886336998473e-05,
+      "loss": 3.8689,
+      "step": 2041856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6560477422474212e-05,
+      "loss": 3.8634,
+      "step": 2042368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.655209147496369e-05,
+      "loss": 3.8623,
+      "step": 2042880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.654370552745317e-05,
+      "loss": 3.8653,
+      "step": 2043392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.653531957994265e-05,
+      "loss": 3.8724,
+      "step": 2043904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.652693363243213e-05,
+      "loss": 3.8621,
+      "step": 2044416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6518564063725337e-05,
+      "loss": 3.8628,
+      "step": 2044928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.651017811621482e-05,
+      "loss": 3.8606,
+      "step": 2045440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6501808547508033e-05,
+      "loss": 3.8682,
+      "step": 2045952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.649342259999751e-05,
+      "loss": 3.8605,
+      "step": 2046464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.648503665248699e-05,
+      "loss": 3.8626,
+      "step": 2046976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.647665070497647e-05,
+      "loss": 3.8611,
+      "step": 2047488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6468281136269682e-05,
+      "loss": 3.8743,
+      "step": 2048000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6459895188759162e-05,
+      "loss": 3.8554,
+      "step": 2048512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6451509241248642e-05,
+      "loss": 3.8564,
+      "step": 2049024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6443123293738122e-05,
+      "loss": 3.8487,
+      "step": 2049536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.643475372503133e-05,
+      "loss": 3.8539,
+      "step": 2050048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.642636777752081e-05,
+      "loss": 3.8703,
+      "step": 2050560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.641798183001029e-05,
+      "loss": 3.8664,
+      "step": 2051072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6409595882499775e-05,
+      "loss": 3.8801,
+      "step": 2051584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6401209934989254e-05,
+      "loss": 3.8516,
+      "step": 2052096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6392823987478734e-05,
+      "loss": 3.8695,
+      "step": 2052608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6384438039968214e-05,
+      "loss": 3.8635,
+      "step": 2053120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.637605209245769e-05,
+      "loss": 3.867,
+      "step": 2053632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.636766614494717e-05,
+      "loss": 3.8591,
+      "step": 2054144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.635928019743665e-05,
+      "loss": 3.8591,
+      "step": 2054656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6350894249926134e-05,
+      "loss": 3.8637,
+      "step": 2055168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6342524681219343e-05,
+      "loss": 3.8783,
+      "step": 2055680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6334138733708823e-05,
+      "loss": 3.8491,
+      "step": 2056192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6325752786198303e-05,
+      "loss": 3.8659,
+      "step": 2056704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6317366838687783e-05,
+      "loss": 3.8677,
+      "step": 2057216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6308980891177263e-05,
+      "loss": 3.8619,
+      "step": 2057728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6300611322470476e-05,
+      "loss": 3.8597,
+      "step": 2058240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6292225374959956e-05,
+      "loss": 3.8663,
+      "step": 2058752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6283839427449436e-05,
+      "loss": 3.8644,
+      "step": 2059264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6275453479938915e-05,
+      "loss": 3.8676,
+      "step": 2059776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6267067532428395e-05,
+      "loss": 3.8668,
+      "step": 2060288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.000288486480713,
+      "eval_runtime": 301.0381,
+      "eval_samples_per_second": 1267.584,
+      "eval_steps_per_second": 39.613,
+      "step": 2060640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6258681584917872e-05,
+      "loss": 3.8559,
+      "step": 2060800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6250295637407352e-05,
+      "loss": 3.8556,
+      "step": 2061312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6241926068700568e-05,
+      "loss": 3.8651,
+      "step": 2061824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6233540121190044e-05,
+      "loss": 3.8655,
+      "step": 2062336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6225170552483257e-05,
+      "loss": 3.8758,
+      "step": 2062848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6216784604972737e-05,
+      "loss": 3.8605,
+      "step": 2063360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6208398657462217e-05,
+      "loss": 3.8682,
+      "step": 2063872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6200012709951697e-05,
+      "loss": 3.8619,
+      "step": 2064384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6191626762441177e-05,
+      "loss": 3.8577,
+      "step": 2064896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6183240814930657e-05,
+      "loss": 3.857,
+      "step": 2065408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.617487124622387e-05,
+      "loss": 3.866,
+      "step": 2065920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6166485298713346e-05,
+      "loss": 3.8666,
+      "step": 2066432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6158115730006558e-05,
+      "loss": 3.8722,
+      "step": 2066944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6149729782496038e-05,
+      "loss": 3.8545,
+      "step": 2067456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6141343834985518e-05,
+      "loss": 3.8558,
+      "step": 2067968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6132957887474998e-05,
+      "loss": 3.8456,
+      "step": 2068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6124571939964478e-05,
+      "loss": 3.8692,
+      "step": 2068992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6116185992453958e-05,
+      "loss": 3.8645,
+      "step": 2069504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6107800044943438e-05,
+      "loss": 3.8543,
+      "step": 2070016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6099414097432918e-05,
+      "loss": 3.8673,
+      "step": 2070528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6091028149922398e-05,
+      "loss": 3.8831,
+      "step": 2071040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6082642202411878e-05,
+      "loss": 3.8702,
+      "step": 2071552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6074256254901358e-05,
+      "loss": 3.8667,
+      "step": 2072064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6065870307390838e-05,
+      "loss": 3.8621,
+      "step": 2072576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6057484359880318e-05,
+      "loss": 3.8699,
+      "step": 2073088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6049098412369798e-05,
+      "loss": 3.8637,
+      "step": 2073600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6040745222466743e-05,
+      "loss": 3.8651,
+      "step": 2074112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6032359274956223e-05,
+      "loss": 3.8641,
+      "step": 2074624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.60239733274457e-05,
+      "loss": 3.862,
+      "step": 2075136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.601558737993518e-05,
+      "loss": 3.8591,
+      "step": 2075648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.600720143242466e-05,
+      "loss": 3.862,
+      "step": 2076160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.599881548491414e-05,
+      "loss": 3.8619,
+      "step": 2076672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.599042953740362e-05,
+      "loss": 3.8679,
+      "step": 2077184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.59820435898931e-05,
+      "loss": 3.8733,
+      "step": 2077696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5973657642382582e-05,
+      "loss": 3.8648,
+      "step": 2078208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.596528807367579e-05,
+      "loss": 3.8606,
+      "step": 2078720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.595690212616527e-05,
+      "loss": 3.8707,
+      "step": 2079232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.594851617865475e-05,
+      "loss": 3.8519,
+      "step": 2079744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.594013023114423e-05,
+      "loss": 3.8572,
+      "step": 2080256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5931744283633708e-05,
+      "loss": 3.8532,
+      "step": 2080768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.592335833612319e-05,
+      "loss": 3.8578,
+      "step": 2081280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.591497238861267e-05,
+      "loss": 3.8635,
+      "step": 2081792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.590660281990588e-05,
+      "loss": 3.8637,
+      "step": 2082304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.589821687239536e-05,
+      "loss": 3.8663,
+      "step": 2082816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5889847303688573e-05,
+      "loss": 3.8608,
+      "step": 2083328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5881461356178053e-05,
+      "loss": 3.8662,
+      "step": 2083840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5873075408667533e-05,
+      "loss": 3.8669,
+      "step": 2084352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5864689461157012e-05,
+      "loss": 3.8739,
+      "step": 2084864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5856303513646492e-05,
+      "loss": 3.8416,
+      "step": 2085376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5847917566135972e-05,
+      "loss": 3.8718,
+      "step": 2085888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5839531618625452e-05,
+      "loss": 3.8584,
+      "step": 2086400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5831145671114932e-05,
+      "loss": 3.8661,
+      "step": 2086912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5822776102408145e-05,
+      "loss": 3.8592,
+      "step": 2087424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5814390154897625e-05,
+      "loss": 3.8562,
+      "step": 2087936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5806004207387105e-05,
+      "loss": 3.8539,
+      "step": 2088448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5797618259876585e-05,
+      "loss": 3.8531,
+      "step": 2088960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.578923231236606e-05,
+      "loss": 3.8765,
+      "step": 2089472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.578084636485554e-05,
+      "loss": 3.8496,
+      "step": 2089984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.577246041734502e-05,
+      "loss": 3.8717,
+      "step": 2090496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.57640744698345e-05,
+      "loss": 3.8591,
+      "step": 2091008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5755688522323984e-05,
+      "loss": 3.851,
+      "step": 2091520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5747318953617194e-05,
+      "loss": 3.8682,
+      "step": 2092032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5738933006106673e-05,
+      "loss": 3.8544,
+      "step": 2092544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5730547058596153e-05,
+      "loss": 3.8443,
+      "step": 2093056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5722161111085633e-05,
+      "loss": 3.8527,
+      "step": 2093568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5713775163575113e-05,
+      "loss": 3.8723,
+      "step": 2094080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5705389216064593e-05,
+      "loss": 3.8536,
+      "step": 2094592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5697019647357806e-05,
+      "loss": 3.8664,
+      "step": 2095104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5688650078651015e-05,
+      "loss": 3.844,
+      "step": 2095616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5680264131140495e-05,
+      "loss": 3.8554,
+      "step": 2096128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5671878183629975e-05,
+      "loss": 3.8623,
+      "step": 2096640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5663508614923187e-05,
+      "loss": 3.8652,
+      "step": 2097152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5655122667412667e-05,
+      "loss": 3.8628,
+      "step": 2097664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5646736719902147e-05,
+      "loss": 3.865,
+      "step": 2098176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5638350772391627e-05,
+      "loss": 3.8679,
+      "step": 2098688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5629964824881107e-05,
+      "loss": 3.8565,
+      "step": 2099200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5621578877370587e-05,
+      "loss": 3.8603,
+      "step": 2099712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.56132093086638e-05,
+      "loss": 3.8544,
+      "step": 2100224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.560482336115328e-05,
+      "loss": 3.8613,
+      "step": 2100736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.559643741364276e-05,
+      "loss": 3.8564,
+      "step": 2101248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.558805146613224e-05,
+      "loss": 3.8578,
+      "step": 2101760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5579665518621716e-05,
+      "loss": 3.8675,
+      "step": 2102272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5571279571111196e-05,
+      "loss": 3.8696,
+      "step": 2102784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5562893623600676e-05,
+      "loss": 3.8588,
+      "step": 2103296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.555450767609016e-05,
+      "loss": 3.8404,
+      "step": 2103808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.554613810738337e-05,
+      "loss": 3.8634,
+      "step": 2104320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5537752159872848e-05,
+      "loss": 3.8632,
+      "step": 2104832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5529366212362328e-05,
+      "loss": 3.8689,
+      "step": 2105344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5520980264851808e-05,
+      "loss": 3.8501,
+      "step": 2105856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5512594317341288e-05,
+      "loss": 3.8643,
+      "step": 2106368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5504208369830768e-05,
+      "loss": 3.8637,
+      "step": 2106880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5495822422320248e-05,
+      "loss": 3.8533,
+      "step": 2107392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.548745285361346e-05,
+      "loss": 3.8528,
+      "step": 2107904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.547906690610294e-05,
+      "loss": 3.8666,
+      "step": 2108416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.547068095859242e-05,
+      "loss": 3.8616,
+      "step": 2108928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5462295011081897e-05,
+      "loss": 3.87,
+      "step": 2109440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5453909063571377e-05,
+      "loss": 3.8545,
+      "step": 2109952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.544552311606086e-05,
+      "loss": 3.8563,
+      "step": 2110464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.543713716855034e-05,
+      "loss": 3.8533,
+      "step": 2110976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.542875122103982e-05,
+      "loss": 3.8661,
+      "step": 2111488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.542038165233303e-05,
+      "loss": 3.8633,
+      "step": 2112000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.541199570482251e-05,
+      "loss": 3.8555,
+      "step": 2112512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.540360975731199e-05,
+      "loss": 3.8601,
+      "step": 2113024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5395256567408934e-05,
+      "loss": 3.8492,
+      "step": 2113536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5386870619898414e-05,
+      "loss": 3.8571,
+      "step": 2114048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5378484672387894e-05,
+      "loss": 3.8426,
+      "step": 2114560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.537009872487737e-05,
+      "loss": 3.8573,
+      "step": 2115072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.536171277736685e-05,
+      "loss": 3.8539,
+      "step": 2115584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.535332682985633e-05,
+      "loss": 3.8622,
+      "step": 2116096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5344940882345814e-05,
+      "loss": 3.8556,
+      "step": 2116608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5336554934835294e-05,
+      "loss": 3.8591,
+      "step": 2117120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5328168987324774e-05,
+      "loss": 3.8484,
+      "step": 2117632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5319799418617983e-05,
+      "loss": 3.866,
+      "step": 2118144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5311413471107463e-05,
+      "loss": 3.8575,
+      "step": 2118656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5303027523596943e-05,
+      "loss": 3.8567,
+      "step": 2119168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5294641576086423e-05,
+      "loss": 3.8597,
+      "step": 2119680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5286255628575903e-05,
+      "loss": 3.8718,
+      "step": 2120192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5277886059869115e-05,
+      "loss": 3.8557,
+      "step": 2120704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5269500112358595e-05,
+      "loss": 3.8577,
+      "step": 2121216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5261114164848075e-05,
+      "loss": 3.8602,
+      "step": 2121728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5252728217337553e-05,
+      "loss": 3.8626,
+      "step": 2122240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5244342269827033e-05,
+      "loss": 3.8563,
+      "step": 2122752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5235956322316513e-05,
+      "loss": 3.8613,
+      "step": 2123264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5227586753609726e-05,
+      "loss": 3.8538,
+      "step": 2123776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5219217184902937e-05,
+      "loss": 3.865,
+      "step": 2124288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5210831237392417e-05,
+      "loss": 3.8484,
+      "step": 2124800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5202445289881898e-05,
+      "loss": 3.8529,
+      "step": 2125312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5194059342371375e-05,
+      "loss": 3.8446,
+      "step": 2125824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5185689773664587e-05,
+      "loss": 3.8546,
+      "step": 2126336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5177303826154069e-05,
+      "loss": 3.8622,
+      "step": 2126848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5168917878643549e-05,
+      "loss": 3.8626,
+      "step": 2127360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5160531931133026e-05,
+      "loss": 3.8713,
+      "step": 2127872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5152145983622507e-05,
+      "loss": 3.8512,
+      "step": 2128384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5143760036111987e-05,
+      "loss": 3.8618,
+      "step": 2128896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5135374088601467e-05,
+      "loss": 3.864,
+      "step": 2129408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5126988141090947e-05,
+      "loss": 3.8627,
+      "step": 2129920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5118602193580429e-05,
+      "loss": 3.8553,
+      "step": 2130432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5110216246069909e-05,
+      "loss": 3.852,
+      "step": 2130944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5101830298559385e-05,
+      "loss": 3.8608,
+      "step": 2131456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5093444351048867e-05,
+      "loss": 3.866,
+      "step": 2131968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5085058403538347e-05,
+      "loss": 3.8518,
+      "step": 2132480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5076672456027827e-05,
+      "loss": 3.8608,
+      "step": 2132992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5068302887321038e-05,
+      "loss": 3.8639,
+      "step": 2133504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5059916939810518e-05,
+      "loss": 3.8584,
+      "step": 2134016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5051530992299997e-05,
+      "loss": 3.8543,
+      "step": 2134528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5043145044789477e-05,
+      "loss": 3.8598,
+      "step": 2135040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5034759097278957e-05,
+      "loss": 3.8585,
+      "step": 2135552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5026389528572168e-05,
+      "loss": 3.8612,
+      "step": 2136064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5018003581061648e-05,
+      "loss": 3.8664,
+      "step": 2136576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9988834857940674,
+      "eval_runtime": 297.0038,
+      "eval_samples_per_second": 1284.802,
+      "eval_steps_per_second": 40.151,
+      "step": 2136960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.500961763355113e-05,
+      "loss": 3.8562,
+      "step": 2137088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.500123168604061e-05,
+      "loss": 3.8474,
+      "step": 2137600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4992845738530086e-05,
+      "loss": 3.8589,
+      "step": 2138112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4984459791019568e-05,
+      "loss": 3.863,
+      "step": 2138624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4976073843509048e-05,
+      "loss": 3.8661,
+      "step": 2139136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4967687895998528e-05,
+      "loss": 3.8594,
+      "step": 2139648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4959301948488008e-05,
+      "loss": 3.8604,
+      "step": 2140160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4950916000977488e-05,
+      "loss": 3.8564,
+      "step": 2140672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.494253005346697e-05,
+      "loss": 3.8519,
+      "step": 2141184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4934144105956446e-05,
+      "loss": 3.8555,
+      "step": 2141696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4925758158445926e-05,
+      "loss": 3.8628,
+      "step": 2142208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4917372210935408e-05,
+      "loss": 3.8616,
+      "step": 2142720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.490900264222862e-05,
+      "loss": 3.8688,
+      "step": 2143232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4900616694718098e-05,
+      "loss": 3.8479,
+      "step": 2143744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4892230747207578e-05,
+      "loss": 3.8505,
+      "step": 2144256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4883844799697058e-05,
+      "loss": 3.8425,
+      "step": 2144768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4875458852186538e-05,
+      "loss": 3.8643,
+      "step": 2145280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4867072904676018e-05,
+      "loss": 3.8597,
+      "step": 2145792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.48586869571655e-05,
+      "loss": 3.857,
+      "step": 2146304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4850301009654976e-05,
+      "loss": 3.8573,
+      "step": 2146816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4841915062144456e-05,
+      "loss": 3.8772,
+      "step": 2147328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.483354549343767e-05,
+      "loss": 3.8671,
+      "step": 2147840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.482517592473088e-05,
+      "loss": 3.8618,
+      "step": 2148352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4816789977220361e-05,
+      "loss": 3.856,
+      "step": 2148864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4808404029709841e-05,
+      "loss": 3.8701,
+      "step": 2149376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4800018082199321e-05,
+      "loss": 3.8578,
+      "step": 2149888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4791632134688801e-05,
+      "loss": 3.8564,
+      "step": 2150400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.478324618717828e-05,
+      "loss": 3.8653,
+      "step": 2150912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4774876618471492e-05,
+      "loss": 3.8533,
+      "step": 2151424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4766490670960972e-05,
+      "loss": 3.8579,
+      "step": 2151936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.475810472345045e-05,
+      "loss": 3.8589,
+      "step": 2152448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4749735154743663e-05,
+      "loss": 3.854,
+      "step": 2152960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4741349207233143e-05,
+      "loss": 3.8615,
+      "step": 2153472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4732963259722624e-05,
+      "loss": 3.8714,
+      "step": 2153984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.47245773122121e-05,
+      "loss": 3.856,
+      "step": 2154496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.471619136470158e-05,
+      "loss": 3.8583,
+      "step": 2155008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4707805417191062e-05,
+      "loss": 3.8659,
+      "step": 2155520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4699419469680542e-05,
+      "loss": 3.8466,
+      "step": 2156032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4691033522170022e-05,
+      "loss": 3.8518,
+      "step": 2156544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4682647574659502e-05,
+      "loss": 3.8544,
+      "step": 2157056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4674261627148982e-05,
+      "loss": 3.8495,
+      "step": 2157568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.466587567963846e-05,
+      "loss": 3.8576,
+      "step": 2158080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.465748973212794e-05,
+      "loss": 3.8606,
+      "step": 2158592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.464910378461742e-05,
+      "loss": 3.8603,
+      "step": 2159104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4640717837106902e-05,
+      "loss": 3.8575,
+      "step": 2159616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4632348268400111e-05,
+      "loss": 3.8577,
+      "step": 2160128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4623962320889593e-05,
+      "loss": 3.8627,
+      "step": 2160640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4615592752182805e-05,
+      "loss": 3.8655,
+      "step": 2161152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4607206804672283e-05,
+      "loss": 3.8422,
+      "step": 2161664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4598820857161763e-05,
+      "loss": 3.8666,
+      "step": 2162176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4590434909651243e-05,
+      "loss": 3.8522,
+      "step": 2162688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4582048962140723e-05,
+      "loss": 3.8598,
+      "step": 2163200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4573663014630203e-05,
+      "loss": 3.8577,
+      "step": 2163712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4565277067119685e-05,
+      "loss": 3.8514,
+      "step": 2164224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4556891119609165e-05,
+      "loss": 3.8497,
+      "step": 2164736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4548505172098641e-05,
+      "loss": 3.8452,
+      "step": 2165248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4540119224588123e-05,
+      "loss": 3.8745,
+      "step": 2165760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4531733277077603e-05,
+      "loss": 3.8461,
+      "step": 2166272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4523347329567083e-05,
+      "loss": 3.8622,
+      "step": 2166784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4514977760860294e-05,
+      "loss": 3.8547,
+      "step": 2167296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4506591813349774e-05,
+      "loss": 3.8489,
+      "step": 2167808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4498205865839254e-05,
+      "loss": 3.8576,
+      "step": 2168320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4489819918328734e-05,
+      "loss": 3.8543,
+      "step": 2168832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4481450349621944e-05,
+      "loss": 3.8414,
+      "step": 2169344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4473097159718888e-05,
+      "loss": 3.8439,
+      "step": 2169856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.44647275910121e-05,
+      "loss": 3.8729,
+      "step": 2170368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4456341643501579e-05,
+      "loss": 3.8486,
+      "step": 2170880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4447955695991059e-05,
+      "loss": 3.8578,
+      "step": 2171392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4439569748480538e-05,
+      "loss": 3.8405,
+      "step": 2171904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4431183800970018e-05,
+      "loss": 3.8522,
+      "step": 2172416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.44227978534595e-05,
+      "loss": 3.8522,
+      "step": 2172928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.441441190594898e-05,
+      "loss": 3.8626,
+      "step": 2173440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.440602595843846e-05,
+      "loss": 3.859,
+      "step": 2173952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4397640010927938e-05,
+      "loss": 3.8613,
+      "step": 2174464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.438927044222115e-05,
+      "loss": 3.8623,
+      "step": 2174976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.438088449471063e-05,
+      "loss": 3.8516,
+      "step": 2175488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4372498547200109e-05,
+      "loss": 3.8539,
+      "step": 2176000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4364112599689589e-05,
+      "loss": 3.8503,
+      "step": 2176512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4355726652179069e-05,
+      "loss": 3.8556,
+      "step": 2177024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4347340704668549e-05,
+      "loss": 3.8547,
+      "step": 2177536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4338954757158029e-05,
+      "loss": 3.8478,
+      "step": 2178048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.433056880964751e-05,
+      "loss": 3.8663,
+      "step": 2178560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.432219924094072e-05,
+      "loss": 3.8655,
+      "step": 2179072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4313813293430201e-05,
+      "loss": 3.8542,
+      "step": 2179584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4305427345919681e-05,
+      "loss": 3.8322,
+      "step": 2180096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4297041398409161e-05,
+      "loss": 3.8572,
+      "step": 2180608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4288671829702372e-05,
+      "loss": 3.859,
+      "step": 2181120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4280285882191852e-05,
+      "loss": 3.8642,
+      "step": 2181632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4271899934681332e-05,
+      "loss": 3.8529,
+      "step": 2182144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4263513987170812e-05,
+      "loss": 3.8573,
+      "step": 2182656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.425512803966029e-05,
+      "loss": 3.8577,
+      "step": 2183168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.424674209214977e-05,
+      "loss": 3.8489,
+      "step": 2183680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4238372523442982e-05,
+      "loss": 3.851,
+      "step": 2184192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4229986575932464e-05,
+      "loss": 3.8619,
+      "step": 2184704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.422160062842194e-05,
+      "loss": 3.8543,
+      "step": 2185216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.421321468091142e-05,
+      "loss": 3.8681,
+      "step": 2185728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4204845112204635e-05,
+      "loss": 3.8534,
+      "step": 2186240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4196459164694115e-05,
+      "loss": 3.85,
+      "step": 2186752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4188073217183593e-05,
+      "loss": 3.846,
+      "step": 2187264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4179687269673073e-05,
+      "loss": 3.8606,
+      "step": 2187776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4171301322162553e-05,
+      "loss": 3.863,
+      "step": 2188288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4162915374652033e-05,
+      "loss": 3.8502,
+      "step": 2188800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4154529427141513e-05,
+      "loss": 3.8542,
+      "step": 2189312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4146143479630994e-05,
+      "loss": 3.8445,
+      "step": 2189824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4137773910924204e-05,
+      "loss": 3.8582,
+      "step": 2190336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4129404342217414e-05,
+      "loss": 3.8335,
+      "step": 2190848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4121018394706894e-05,
+      "loss": 3.8489,
+      "step": 2191360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4112632447196374e-05,
+      "loss": 3.8554,
+      "step": 2191872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4104246499685856e-05,
+      "loss": 3.8583,
+      "step": 2192384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4095860552175336e-05,
+      "loss": 3.8464,
+      "step": 2192896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4087474604664816e-05,
+      "loss": 3.8539,
+      "step": 2193408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4079088657154296e-05,
+      "loss": 3.8446,
+      "step": 2193920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4070719088447507e-05,
+      "loss": 3.8606,
+      "step": 2194432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4062333140936987e-05,
+      "loss": 3.8535,
+      "step": 2194944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4053947193426466e-05,
+      "loss": 3.8533,
+      "step": 2195456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4045561245915945e-05,
+      "loss": 3.8525,
+      "step": 2195968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4037175298405425e-05,
+      "loss": 3.8685,
+      "step": 2196480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4028805729698639e-05,
+      "loss": 3.8522,
+      "step": 2196992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.402045253979558e-05,
+      "loss": 3.8537,
+      "step": 2197504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4012066592285062e-05,
+      "loss": 3.857,
+      "step": 2198016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4003680644774539e-05,
+      "loss": 3.8553,
+      "step": 2198528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3995294697264019e-05,
+      "loss": 3.8513,
+      "step": 2199040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.39869087497535e-05,
+      "loss": 3.8569,
+      "step": 2199552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.397852280224298e-05,
+      "loss": 3.8497,
+      "step": 2200064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.397013685473246e-05,
+      "loss": 3.861,
+      "step": 2200576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.396175090722194e-05,
+      "loss": 3.8429,
+      "step": 2201088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3953364959711418e-05,
+      "loss": 3.8528,
+      "step": 2201600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3944979012200898e-05,
+      "loss": 3.842,
+      "step": 2202112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3936593064690378e-05,
+      "loss": 3.8523,
+      "step": 2202624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3928207117179858e-05,
+      "loss": 3.8517,
+      "step": 2203136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.391982116966934e-05,
+      "loss": 3.8591,
+      "step": 2203648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.391143522215882e-05,
+      "loss": 3.8678,
+      "step": 2204160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.39030492746483e-05,
+      "loss": 3.8472,
+      "step": 2204672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.389467970594151e-05,
+      "loss": 3.858,
+      "step": 2205184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.388629375843099e-05,
+      "loss": 3.8567,
+      "step": 2205696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.387790781092047e-05,
+      "loss": 3.857,
+      "step": 2206208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3869538242213681e-05,
+      "loss": 3.8549,
+      "step": 2206720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3861152294703161e-05,
+      "loss": 3.8441,
+      "step": 2207232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3852766347192641e-05,
+      "loss": 3.8583,
+      "step": 2207744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3844380399682121e-05,
+      "loss": 3.8631,
+      "step": 2208256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.38359944521716e-05,
+      "loss": 3.8472,
+      "step": 2208768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.382760850466108e-05,
+      "loss": 3.8531,
+      "step": 2209280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3819238935954294e-05,
+      "loss": 3.8625,
+      "step": 2209792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3810852988443774e-05,
+      "loss": 3.8521,
+      "step": 2210304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.380246704093325e-05,
+      "loss": 3.8506,
+      "step": 2210816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3794081093422732e-05,
+      "loss": 3.8597,
+      "step": 2211328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3785711524715944e-05,
+      "loss": 3.8533,
+      "step": 2211840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3777341956009155e-05,
+      "loss": 3.8541,
+      "step": 2212352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3768956008498635e-05,
+      "loss": 3.8646,
+      "step": 2212864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.997385263442993,
+      "eval_runtime": 296.3881,
+      "eval_samples_per_second": 1287.471,
+      "eval_steps_per_second": 40.234,
+      "step": 2213280
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 8.862269296767544e+17,
+  "trial_name": null,
+  "trial_params": null
+}