diff --git "a/checkpoint-2518560/trainer_state.json" "b/checkpoint-2518560/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2518560/trainer_state.json"
@@ -0,0 +1,29803 @@
+{
+  "best_metric": 3.973715305328369,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-sent-neg/lstm/3/checkpoints/checkpoint-2518560",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 2518560,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8187,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5487,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0597,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9956,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9491,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9236,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7673,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6605,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5615,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4811,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.4166,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3434,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2739,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1955,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1487,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0869,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.044,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9981,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9589,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.917,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8838,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8415,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8082,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7764,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7561,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7185,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6943,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6664,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6364,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6261,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6005,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 5.5925,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.556,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5449,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5334,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5162,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.4948,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.46,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9681399509815164e-05,
+      "loss": 5.4632,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9673013562304644e-05,
+      "loss": 5.429,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4315,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4151,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4045,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.3846,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3785,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3782,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3608,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3561,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3276,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3134,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3225,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3091,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3008,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2703,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.286,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2603,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2814,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2324,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2438,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2241,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2122,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2144,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2016,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.1798,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.1807,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.2013,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 5.1745,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 5.1634,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 5.152,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 5.1378,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.1515,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1429,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1357,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1418,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1269,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1211,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.0985,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.099,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.0898,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0805,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0804,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0785,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0806,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0692,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0437,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0544,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0547,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0554,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0337,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.037,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0288,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0207,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0154,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0115,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 4.9989,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 5.0067,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 4.9898,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9809,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.9775,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9819,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9759,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9611,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9623,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.9493,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9535,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9288,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9392,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9317,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9285,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9313,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.9255,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.904,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.913,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.9153,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.9072,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.907,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.8983,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.8945,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.8826,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.8892,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.8845,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8695,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8716,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8665,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8736,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8501,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8564,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8429,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.851,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8394,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8539,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8413,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8344,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.833,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8279,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8365,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8333,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8244,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8187,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8202,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8096,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.7973,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8054,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.7966,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.7888,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.79,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.7877,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.7912,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7911,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7787,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.7453789710998535,
+      "eval_runtime": 301.8396,
+      "eval_samples_per_second": 1264.218,
+      "eval_steps_per_second": 39.508,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.769,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7668,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7781,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7697,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7717,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7534,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7553,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7501,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7553,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667092952331785e-05,
+      "loss": 4.754,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7511,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865032105731075e-05,
+      "loss": 4.7574,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7257,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7374,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7316,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7222,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7355,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7259,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.7222,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7441,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857489666612726e-05,
+      "loss": 4.7242,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.7179,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7171,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.7316,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7105,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.7052,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7058,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7019,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7012,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.6945,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.7137,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.6941,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.6929,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.6948,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.6935,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.6874,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6686,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6766,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6618,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6796,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.6675,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.6728,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6631,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838208538860022e-05,
+      "loss": 4.6676,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6705,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.663,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.6663,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.6531,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834020478745882e-05,
+      "loss": 4.6455,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331818839948295e-05,
+      "loss": 4.6571,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.6515,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.6565,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.6289,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.6521,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289905481199424e-05,
+      "loss": 4.6297,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.6618,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273133586178384e-05,
+      "loss": 4.6242,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.6398,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.6291,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824797574364683e-05,
+      "loss": 4.6216,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.6257,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.6241,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.6067,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.6132,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820606238489796e-05,
+      "loss": 4.6288,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819767643738744e-05,
+      "loss": 4.6155,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.6117,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6075,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.5979,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6164,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.6106,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.6116,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.6166,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.6133,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.6115,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.5901,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.6017,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809707782486866e-05,
+      "loss": 4.5959,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.5843,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808032230865135e-05,
+      "loss": 4.5946,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.5944,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6033,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.5956,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046794897413e-05,
+      "loss": 4.5747,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803840894990248e-05,
+      "loss": 4.5844,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803002300239196e-05,
+      "loss": 4.5929,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.5948,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.5843,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.5896,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.5851,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.5725,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.5782,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797133774862205e-05,
+      "loss": 4.5791,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796295180111153e-05,
+      "loss": 4.5686,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.5834,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5649,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793781033738371e-05,
+      "loss": 4.5621,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792942438987319e-05,
+      "loss": 4.5651,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.5721,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.5693,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5579,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.5561,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5512,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.5619,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870739136103276e-05,
+      "loss": 4.5383,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862369567396485e-05,
+      "loss": 4.5467,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853983619885965e-05,
+      "loss": 4.5543,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845597672375445e-05,
+      "loss": 4.5489,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837211724864925e-05,
+      "loss": 4.5484,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.5472,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782045620864762e-05,
+      "loss": 4.5365,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78120702611371e-05,
+      "loss": 4.5444,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.5444,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.5434,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.5474,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.5466,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777015690238823e-05,
+      "loss": 4.5378,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.5318,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.5394,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5344,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.5235,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.5318,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.5267,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.5355,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77030857011078e-05,
+      "loss": 4.518,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.5272,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768631380608676e-05,
+      "loss": 4.5064,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.5284,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.5149,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.532,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5243,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.5194,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636030878631096e-05,
+      "loss": 4.5112,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762764493112058e-05,
+      "loss": 4.513,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761925898361006e-05,
+      "loss": 4.5269,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.5295,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760251984619649e-05,
+      "loss": 4.5181,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759413389868596e-05,
+      "loss": 4.5144,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758574795117544e-05,
+      "loss": 4.5117,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757736200366492e-05,
+      "loss": 4.5085,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75689760561544e-05,
+      "loss": 4.5035,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756059010864388e-05,
+      "loss": 4.5093,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755220416113336e-05,
+      "loss": 4.5035,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754381821362284e-05,
+      "loss": 4.4966,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753543226611232e-05,
+      "loss": 4.4994,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75270463186018e-05,
+      "loss": 4.5001,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751866037109128e-05,
+      "loss": 4.5022,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510290802384496e-05,
+      "loss": 4.5016,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501904854873976e-05,
+      "loss": 4.4985,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.4658379554748535,
+      "eval_runtime": 298.8499,
+      "eval_samples_per_second": 1276.865,
+      "eval_steps_per_second": 39.903,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493518907363456e-05,
+      "loss": 4.4838,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485132959852936e-05,
+      "loss": 4.4883,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476747012342416e-05,
+      "loss": 4.4969,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468361064831896e-05,
+      "loss": 4.4946,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459975117321376e-05,
+      "loss": 4.4977,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745158916981085e-05,
+      "loss": 4.4775,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744320322230033e-05,
+      "loss": 4.4889,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743481727478981e-05,
+      "loss": 4.4785,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742643132727929e-05,
+      "loss": 4.4889,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741804537976877e-05,
+      "loss": 4.4876,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740965943225825e-05,
+      "loss": 4.4873,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401273484747736e-05,
+      "loss": 4.4901,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392887537237216e-05,
+      "loss": 4.4648,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384501589726696e-05,
+      "loss": 4.4752,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376115642216176e-05,
+      "loss": 4.4756,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367729694705656e-05,
+      "loss": 4.467,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359343747195135e-05,
+      "loss": 4.4763,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350957799684615e-05,
+      "loss": 4.47,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4703,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334202283467305e-05,
+      "loss": 4.4981,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325816335956784e-05,
+      "loss": 4.4722,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317430388446264e-05,
+      "loss": 4.4733,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730904444093574e-05,
+      "loss": 4.4692,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300674872228953e-05,
+      "loss": 4.4853,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292288924718433e-05,
+      "loss": 4.4646,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728390297720791e-05,
+      "loss": 4.4623,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727551702969739e-05,
+      "loss": 4.4663,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726714746099061e-05,
+      "loss": 4.4662,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725876151348009e-05,
+      "loss": 4.4581,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725037556596957e-05,
+      "loss": 4.4605,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724200599726278e-05,
+      "loss": 4.4758,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723362004975226e-05,
+      "loss": 4.4616,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722523410224174e-05,
+      "loss": 4.4582,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721684815473121e-05,
+      "loss": 4.4654,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.4608,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.457,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719170669100339e-05,
+      "loss": 4.4395,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718332074349287e-05,
+      "loss": 4.4556,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717493479598235e-05,
+      "loss": 4.4386,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716654884847183e-05,
+      "loss": 4.4521,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715816290096131e-05,
+      "loss": 4.4452,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714977695345079e-05,
+      "loss": 4.4538,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714139100594027e-05,
+      "loss": 4.4448,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713300505842975e-05,
+      "loss": 4.4455,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712466824733042e-05,
+      "loss": 4.4497,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71162822998199e-05,
+      "loss": 4.4482,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710789635230938e-05,
+      "loss": 4.4512,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709951040479886e-05,
+      "loss": 4.4372,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709112445728834e-05,
+      "loss": 4.4283,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708273850977782e-05,
+      "loss": 4.4476,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70743525622673e-05,
+      "loss": 4.4383,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706596661475678e-05,
+      "loss": 4.4424,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705758066724626e-05,
+      "loss": 4.4218,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704921109853947e-05,
+      "loss": 4.4458,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704082515102895e-05,
+      "loss": 4.4213,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703243920351843e-05,
+      "loss": 4.4509,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702405325600791e-05,
+      "loss": 4.4141,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701566730849739e-05,
+      "loss": 4.4368,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700728136098687e-05,
+      "loss": 4.4252,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699889541347635e-05,
+      "loss": 4.4238,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699050946596583e-05,
+      "loss": 4.4233,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982139897259045e-05,
+      "loss": 4.4242,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973770328552254e-05,
+      "loss": 4.4066,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965384381041734e-05,
+      "loss": 4.4146,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956998433531214e-05,
+      "loss": 4.426,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948612486020694e-05,
+      "loss": 4.4234,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940226538510174e-05,
+      "loss": 4.4131,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931840590999654e-05,
+      "loss": 4.4139,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923454643489134e-05,
+      "loss": 4.4102,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915068695978614e-05,
+      "loss": 4.4169,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690671550607555e-05,
+      "loss": 4.4198,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689832955856503e-05,
+      "loss": 4.419,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688994361105451e-05,
+      "loss": 4.4277,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881557663544e-05,
+      "loss": 4.4275,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687317171603348e-05,
+      "loss": 4.4221,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686478576852296e-05,
+      "loss": 4.4032,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685641619981617e-05,
+      "loss": 4.4131,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684803025230565e-05,
+      "loss": 4.4168,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683966068359886e-05,
+      "loss": 4.3985,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683127473608834e-05,
+      "loss": 4.4122,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682288878857782e-05,
+      "loss": 4.4083,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68145028410673e-05,
+      "loss": 4.4191,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680611689355678e-05,
+      "loss": 4.4124,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679773094604626e-05,
+      "loss": 4.3936,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678934499853574e-05,
+      "loss": 4.4071,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678095905102522e-05,
+      "loss": 4.4109,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772573103514697e-05,
+      "loss": 4.4145,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764187156004177e-05,
+      "loss": 4.4071,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755801208493657e-05,
+      "loss": 4.4117,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674743163978687e-05,
+      "loss": 4.4093,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739045692276346e-05,
+      "loss": 4.3943,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730659744765826e-05,
+      "loss": 4.4063,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722273797255305e-05,
+      "loss": 4.4017,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713887849744785e-05,
+      "loss": 4.3941,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705501902234265e-05,
+      "loss": 4.4133,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697132333527474e-05,
+      "loss": 4.3952,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688746386016954e-05,
+      "loss": 4.3861,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680360438506434e-05,
+      "loss": 4.395,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667197449099592e-05,
+      "loss": 4.4033,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66635885434854e-05,
+      "loss": 4.3996,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665520259597488e-05,
+      "loss": 4.3901,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664683302726809e-05,
+      "loss": 4.3853,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66384634585613e-05,
+      "loss": 4.3837,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663007751105078e-05,
+      "loss": 4.395,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662169156354026e-05,
+      "loss": 4.3729,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661330561602974e-05,
+      "loss": 4.3846,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660491966851922e-05,
+      "loss": 4.3892,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65965337210087e-05,
+      "loss": 4.3859,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658814777349818e-05,
+      "loss": 4.3841,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657976182598766e-05,
+      "loss": 4.3872,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657139225728087e-05,
+      "loss": 4.3774,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563006309770355e-05,
+      "loss": 4.383,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554620362259835e-05,
+      "loss": 4.3828,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546234414749315e-05,
+      "loss": 4.381,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537848467238795e-05,
+      "loss": 4.3888,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529462519728275e-05,
+      "loss": 4.3883,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521092951021484e-05,
+      "loss": 4.3857,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512707003510964e-05,
+      "loss": 4.3715,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504321056000444e-05,
+      "loss": 4.3788,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495935108489924e-05,
+      "loss": 4.377,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487549160979403e-05,
+      "loss": 4.3668,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479163213468883e-05,
+      "loss": 4.3815,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647077726595836e-05,
+      "loss": 4.3682,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462391318447837e-05,
+      "loss": 4.3817,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645402174974105e-05,
+      "loss": 4.3651,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644563580223053e-05,
+      "loss": 4.3752,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643724985472001e-05,
+      "loss": 4.3565,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642888028601323e-05,
+      "loss": 4.3757,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642049433850271e-05,
+      "loss": 4.3608,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641210839099218e-05,
+      "loss": 4.3817,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640372244348166e-05,
+      "loss": 4.3793,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639533649597114e-05,
+      "loss": 4.3692,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638695054846062e-05,
+      "loss": 4.3615,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637858097975383e-05,
+      "loss": 4.3657,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637019503224331e-05,
+      "loss": 4.3773,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636180908473279e-05,
+      "loss": 4.3835,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635342313722228e-05,
+      "loss": 4.3683,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634503718971176e-05,
+      "loss": 4.3697,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336667621004966e-05,
+      "loss": 4.3675,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328281673494446e-05,
+      "loss": 4.3656,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319895725983926e-05,
+      "loss": 4.3574,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311509778473406e-05,
+      "loss": 4.3613,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303123830962886e-05,
+      "loss": 4.3629,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294737883452366e-05,
+      "loss": 4.3542,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286368314745575e-05,
+      "loss": 4.3553,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277982367235055e-05,
+      "loss": 4.3608,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269596419724535e-05,
+      "loss": 4.3543,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261210472214015e-05,
+      "loss": 4.3613,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625284090350723e-05,
+      "loss": 4.3554,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.332054615020752,
+      "eval_runtime": 293.744,
+      "eval_samples_per_second": 1299.06,
+      "eval_steps_per_second": 40.597,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624445495599671e-05,
+      "loss": 4.3516,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623606900848619e-05,
+      "loss": 4.3474,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622768306097567e-05,
+      "loss": 4.3614,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621929711346515e-05,
+      "loss": 4.3507,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621091116595463e-05,
+      "loss": 4.3623,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620252521844411e-05,
+      "loss": 4.3395,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619415564973732e-05,
+      "loss": 4.3517,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61857697022268e-05,
+      "loss": 4.3396,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617738375471628e-05,
+      "loss": 4.3559,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616899780720576e-05,
+      "loss": 4.3505,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616061185969524e-05,
+      "loss": 4.3525,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615222591218472e-05,
+      "loss": 4.3555,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61438399646742e-05,
+      "loss": 4.3321,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613545401716368e-05,
+      "loss": 4.3415,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612706806965316e-05,
+      "loss": 4.3408,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611868212214264e-05,
+      "loss": 4.3385,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611029617463212e-05,
+      "loss": 4.34,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61019102271216e-05,
+      "loss": 4.344,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609354065841481e-05,
+      "loss": 4.3352,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608515471090429e-05,
+      "loss": 4.3694,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607676876339377e-05,
+      "loss": 4.3391,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606838281588325e-05,
+      "loss": 4.3419,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606001324717646e-05,
+      "loss": 4.341,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605162729966594e-05,
+      "loss": 4.3519,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604324135215542e-05,
+      "loss": 4.3335,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603487178344863e-05,
+      "loss": 4.3399,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602648583593811e-05,
+      "loss": 4.3341,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601809988842759e-05,
+      "loss": 4.3372,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600971394091707e-05,
+      "loss": 4.3267,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600132799340655e-05,
+      "loss": 4.3366,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599294204589603e-05,
+      "loss": 4.3461,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598455609838551e-05,
+      "loss": 4.3384,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597617015087499e-05,
+      "loss": 4.3365,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59678005821682e-05,
+      "loss": 4.3379,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595943101346141e-05,
+      "loss": 4.3328,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595104506595089e-05,
+      "loss": 4.3369,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594265911844037e-05,
+      "loss": 4.3157,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593427317092985e-05,
+      "loss": 4.3322,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592588722341933e-05,
+      "loss": 4.3127,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917517654712546e-05,
+      "loss": 4.3311,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909131707202026e-05,
+      "loss": 4.3241,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900745759691506e-05,
+      "loss": 4.3317,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892359812180986e-05,
+      "loss": 4.3218,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883973864670466e-05,
+      "loss": 4.3224,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875604295963675e-05,
+      "loss": 4.3307,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867218348453155e-05,
+      "loss": 4.3271,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858832400942635e-05,
+      "loss": 4.3273,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850446453432115e-05,
+      "loss": 4.3218,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842060505921595e-05,
+      "loss": 4.3066,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833690937214804e-05,
+      "loss": 4.3287,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825304989704284e-05,
+      "loss": 4.3196,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581691904219377e-05,
+      "loss": 4.3233,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580853309468325e-05,
+      "loss": 4.308,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580014714717273e-05,
+      "loss": 4.3311,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791761199662204e-05,
+      "loss": 4.2995,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783375252151684e-05,
+      "loss": 4.3349,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774989304641164e-05,
+      "loss": 4.2988,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576661973593438e-05,
+      "loss": 4.3159,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575823378842385e-05,
+      "loss": 4.3113,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574984784091333e-05,
+      "loss": 4.3091,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574147827220655e-05,
+      "loss": 4.3043,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573309232469602e-05,
+      "loss": 4.3162,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572470637718551e-05,
+      "loss": 4.2906,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571632042967499e-05,
+      "loss": 4.2985,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570793448216447e-05,
+      "loss": 4.314,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569954853465395e-05,
+      "loss": 4.309,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569116258714343e-05,
+      "loss": 4.3022,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568277663963291e-05,
+      "loss": 4.302,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567440707092612e-05,
+      "loss": 4.2917,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566603750221933e-05,
+      "loss": 4.3053,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565765155470881e-05,
+      "loss": 4.3103,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564926560719829e-05,
+      "loss": 4.3085,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640879659687767e-05,
+      "loss": 4.3184,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632493712177247e-05,
+      "loss": 4.3155,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562412414347046e-05,
+      "loss": 4.3132,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561573819595994e-05,
+      "loss": 4.2934,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560735224844942e-05,
+      "loss": 4.3012,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55989663009389e-05,
+      "loss": 4.308,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559059673223211e-05,
+      "loss": 4.2893,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558221078472159e-05,
+      "loss": 4.3059,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557382483721107e-05,
+      "loss": 4.2988,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556543888970055e-05,
+      "loss": 4.3078,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555705294219003e-05,
+      "loss": 4.305,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554866699467951e-05,
+      "loss": 4.2876,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554028104716899e-05,
+      "loss": 4.3023,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553189509965847e-05,
+      "loss": 4.3007,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552352553095168e-05,
+      "loss": 4.31,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551513958344116e-05,
+      "loss": 4.303,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550675363593064e-05,
+      "loss": 4.3031,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549836768842013e-05,
+      "loss": 4.3067,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489998119713336e-05,
+      "loss": 4.2881,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481612172202816e-05,
+      "loss": 4.3027,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473226224692296e-05,
+      "loss": 4.2943,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464840277181776e-05,
+      "loss": 4.2949,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456470708474985e-05,
+      "loss": 4.3076,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448084760964465e-05,
+      "loss": 4.2953,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439698813453945e-05,
+      "loss": 4.2811,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431312865943425e-05,
+      "loss": 4.2919,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422943297236634e-05,
+      "loss": 4.3013,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414557349726114e-05,
+      "loss": 4.2982,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406171402215594e-05,
+      "loss": 4.2886,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539778545470508e-05,
+      "loss": 4.2836,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538941588599829e-05,
+      "loss": 4.2838,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538102993848777e-05,
+      "loss": 4.2967,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537264399097725e-05,
+      "loss": 4.2692,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536425804346673e-05,
+      "loss": 4.2847,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535588847475994e-05,
+      "loss": 4.2884,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534750252724942e-05,
+      "loss": 4.2878,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53391165797389e-05,
+      "loss": 4.2807,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533073063222838e-05,
+      "loss": 4.2902,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532237744232532e-05,
+      "loss": 4.2808,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53139914948148e-05,
+      "loss": 4.2809,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530560554730428e-05,
+      "loss": 4.2865,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529721959979376e-05,
+      "loss": 4.2815,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528883365228324e-05,
+      "loss": 4.2957,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528044770477272e-05,
+      "loss": 4.2868,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527207813606593e-05,
+      "loss": 4.2882,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526369218855541e-05,
+      "loss": 4.273,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525530624104489e-05,
+      "loss": 4.2827,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524692029353437e-05,
+      "loss": 4.2818,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523853434602385e-05,
+      "loss": 4.2704,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523014839851333e-05,
+      "loss": 4.2836,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221762451002805e-05,
+      "loss": 4.2738,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213376503492285e-05,
+      "loss": 4.2853,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520499055598177e-05,
+      "loss": 4.2703,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519660460847125e-05,
+      "loss": 4.2802,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518821866096073e-05,
+      "loss": 4.256,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517983271345021e-05,
+      "loss": 4.2838,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517146314474342e-05,
+      "loss": 4.2687,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51630771972329e-05,
+      "loss": 4.2836,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515470762852611e-05,
+      "loss": 4.282,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146338059819326e-05,
+      "loss": 4.281,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137952112308806e-05,
+      "loss": 4.2646,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512956616479828e-05,
+      "loss": 4.2762,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512118021728776e-05,
+      "loss": 4.2815,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511279426977724e-05,
+      "loss": 4.2868,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104408322266726e-05,
+      "loss": 4.2777,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096038753559935e-05,
+      "loss": 4.2759,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087652806049415e-05,
+      "loss": 4.2756,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079266858538895e-05,
+      "loss": 4.2789,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070880911028375e-05,
+      "loss": 4.2614,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062494963517855e-05,
+      "loss": 4.2702,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054109016007334e-05,
+      "loss": 4.2695,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045723068496814e-05,
+      "loss": 4.2649,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037337120986294e-05,
+      "loss": 4.2626,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028951173475774e-05,
+      "loss": 4.2722,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020581604768983e-05,
+      "loss": 4.2609,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501219565725846e-05,
+      "loss": 4.269,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500380970974794e-05,
+      "loss": 4.2729,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.2502336502075195,
+      "eval_runtime": 294.515,
+      "eval_samples_per_second": 1295.659,
+      "eval_steps_per_second": 40.49,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499542376223742e-05,
+      "loss": 4.2551,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498703781472691e-05,
+      "loss": 4.2585,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497865186721639e-05,
+      "loss": 4.27,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497026591970587e-05,
+      "loss": 4.2617,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496187997219535e-05,
+      "loss": 4.2749,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495349402468482e-05,
+      "loss": 4.2494,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49451080771743e-05,
+      "loss": 4.2602,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493672212966378e-05,
+      "loss": 4.2522,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492833618215326e-05,
+      "loss": 4.2662,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491995023464274e-05,
+      "loss": 4.2655,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491156428713222e-05,
+      "loss": 4.2568,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49031783396217e-05,
+      "loss": 4.2727,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489479239211118e-05,
+      "loss": 4.241,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488640644460066e-05,
+      "loss": 4.2544,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487803687589387e-05,
+      "loss": 4.2584,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486965092838336e-05,
+      "loss": 4.248,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486126498087284e-05,
+      "loss": 4.2534,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485287903336232e-05,
+      "loss": 4.2564,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48444930858518e-05,
+      "loss": 4.2515,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483610713834128e-05,
+      "loss": 4.2806,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482772119083076e-05,
+      "loss": 4.255,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481933524332024e-05,
+      "loss": 4.2536,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481094929580971e-05,
+      "loss": 4.2611,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480257972710293e-05,
+      "loss": 4.268,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479419377959241e-05,
+      "loss": 4.2483,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478580783208188e-05,
+      "loss": 4.2584,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477742188457136e-05,
+      "loss": 4.2497,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476903593706084e-05,
+      "loss": 4.2527,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476064998955033e-05,
+      "loss": 4.2447,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475226404203981e-05,
+      "loss": 4.2532,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474387809452929e-05,
+      "loss": 4.2643,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4735508525822496e-05,
+      "loss": 4.2538,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727122578311976e-05,
+      "loss": 4.2546,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718736630801456e-05,
+      "loss": 4.2565,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710367062094665e-05,
+      "loss": 4.251,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701981114584145e-05,
+      "loss": 4.2559,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693595167073625e-05,
+      "loss": 4.2345,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685209219563105e-05,
+      "loss": 4.2476,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676823272052585e-05,
+      "loss": 4.2368,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668437324542065e-05,
+      "loss": 4.2433,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466006775583528e-05,
+      "loss": 4.2454,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465168180832476e-05,
+      "loss": 4.2521,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464329586081424e-05,
+      "loss": 4.2408,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463490991330372e-05,
+      "loss": 4.2437,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46265239657932e-05,
+      "loss": 4.248,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461813801828268e-05,
+      "loss": 4.2463,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460976844957589e-05,
+      "loss": 4.247,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460138250206537e-05,
+      "loss": 4.2424,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459299655455485e-05,
+      "loss": 4.2262,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458461060704433e-05,
+      "loss": 4.2501,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457624103833754e-05,
+      "loss": 4.2367,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456785509082702e-05,
+      "loss": 4.2475,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45594691433165e-05,
+      "loss": 4.2329,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455108319580598e-05,
+      "loss": 4.2469,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542697248295465e-05,
+      "loss": 4.2234,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534311300784945e-05,
+      "loss": 4.254,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525925353274425e-05,
+      "loss": 4.221,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45175394057639e-05,
+      "loss": 4.2317,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450918621586084e-05,
+      "loss": 4.2383,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450080026835032e-05,
+      "loss": 4.2299,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44924143208398e-05,
+      "loss": 4.2315,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448402837332928e-05,
+      "loss": 4.2351,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447564242581876e-05,
+      "loss": 4.2225,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446725647830824e-05,
+      "loss": 4.2193,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445888690960145e-05,
+      "loss": 4.2332,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445050096209093e-05,
+      "loss": 4.2354,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444211501458041e-05,
+      "loss": 4.2235,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44337290670699e-05,
+      "loss": 4.2319,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442535949836311e-05,
+      "loss": 4.2138,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441697355085259e-05,
+      "loss": 4.23,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440858760334207e-05,
+      "loss": 4.2337,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440020165583154e-05,
+      "loss": 4.2309,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439181570832102e-05,
+      "loss": 4.2407,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43834297608105e-05,
+      "loss": 4.2382,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437504381329998e-05,
+      "loss": 4.2439,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4366690623396926e-05,
+      "loss": 4.2186,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4358304675886406e-05,
+      "loss": 4.2269,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4349918728375886e-05,
+      "loss": 4.2338,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341532780865366e-05,
+      "loss": 4.2121,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333146833354846e-05,
+      "loss": 4.2347,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324760885844326e-05,
+      "loss": 4.2265,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431639131713754e-05,
+      "loss": 4.2331,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308005369627015e-05,
+      "loss": 4.2329,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299619422116495e-05,
+      "loss": 4.2141,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291233474605975e-05,
+      "loss": 4.2292,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282847527095455e-05,
+      "loss": 4.2255,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274461579584935e-05,
+      "loss": 4.2336,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266075632074415e-05,
+      "loss": 4.2281,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257689684563894e-05,
+      "loss": 4.2346,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249320115857104e-05,
+      "loss": 4.2304,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424095054715032e-05,
+      "loss": 4.2159,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42325645996398e-05,
+      "loss": 4.2286,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422417865212928e-05,
+      "loss": 4.2192,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421579270461876e-05,
+      "loss": 4.2269,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420740675710824e-05,
+      "loss": 4.2348,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419902080959772e-05,
+      "loss": 4.2185,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41906348620872e-05,
+      "loss": 4.214,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418224891457668e-05,
+      "loss": 4.2183,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417386296706616e-05,
+      "loss": 4.2276,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416549339835937e-05,
+      "loss": 4.2316,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415710745084885e-05,
+      "loss": 4.2156,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414872150333833e-05,
+      "loss": 4.215,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414033555582781e-05,
+      "loss": 4.2112,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4131965987121024e-05,
+      "loss": 4.2263,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123580039610504e-05,
+      "loss": 4.2002,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115194092099984e-05,
+      "loss": 4.2138,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106808144589464e-05,
+      "loss": 4.2188,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098422197078944e-05,
+      "loss": 4.2143,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4090036249568424e-05,
+      "loss": 4.2121,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081650302057904e-05,
+      "loss": 4.2233,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407326435454738e-05,
+      "loss": 4.2112,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406487840703686e-05,
+      "loss": 4.2076,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405650883833007e-05,
+      "loss": 4.2194,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404812289081955e-05,
+      "loss": 4.2152,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039736943309026e-05,
+      "loss": 4.2248,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403135099579851e-05,
+      "loss": 4.2198,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402298142709173e-05,
+      "loss": 4.2184,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40145954795812e-05,
+      "loss": 4.2034,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400620953207068e-05,
+      "loss": 4.2134,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399782358456016e-05,
+      "loss": 4.213,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398943763704964e-05,
+      "loss": 4.2003,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398105168953912e-05,
+      "loss": 4.2201,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397268212083233e-05,
+      "loss": 4.2026,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396429617332181e-05,
+      "loss": 4.2169,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395591022581129e-05,
+      "loss": 4.2065,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394752427830077e-05,
+      "loss": 4.2118,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393915470959398e-05,
+      "loss": 4.191,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930768762083466e-05,
+      "loss": 4.2107,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922382814572946e-05,
+      "loss": 4.2067,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3913996867062426e-05,
+      "loss": 4.2118,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905610919551906e-05,
+      "loss": 4.2141,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897224972041386e-05,
+      "loss": 4.2192,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888839024530866e-05,
+      "loss": 4.1961,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880469455824075e-05,
+      "loss": 4.2117,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872083508313555e-05,
+      "loss": 4.2176,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863697560803035e-05,
+      "loss": 4.2172,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855311613292515e-05,
+      "loss": 4.2123,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846925665781995e-05,
+      "loss": 4.214,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838539718271475e-05,
+      "loss": 4.2082,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830153770760955e-05,
+      "loss": 4.2125,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821767823250435e-05,
+      "loss": 4.1936,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813398254543644e-05,
+      "loss": 4.2071,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380501230703313e-05,
+      "loss": 4.2013,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379662635952261e-05,
+      "loss": 4.2031,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378825679081582e-05,
+      "loss": 4.2013,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37798708433053e-05,
+      "loss": 4.2026,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377148489579478e-05,
+      "loss": 4.2021,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376309894828426e-05,
+      "loss": 4.2003,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375471300077374e-05,
+      "loss": 4.2103,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.194133281707764,
+      "eval_runtime": 312.5876,
+      "eval_samples_per_second": 1220.749,
+      "eval_steps_per_second": 38.149,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374634343206695e-05,
+      "loss": 4.1983,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373795748455643e-05,
+      "loss": 4.1925,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372957153704591e-05,
+      "loss": 4.2049,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372120196833912e-05,
+      "loss": 4.2017,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37128160208286e-05,
+      "loss": 4.2041,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3704430073318084e-05,
+      "loss": 4.1937,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3696044125807564e-05,
+      "loss": 4.1923,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368765817829704e-05,
+      "loss": 4.193,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367928860959025e-05,
+      "loss": 4.1993,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367090266207973e-05,
+      "loss": 4.2076,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366251671456921e-05,
+      "loss": 4.1943,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654130767058686e-05,
+      "loss": 4.2074,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645744819548166e-05,
+      "loss": 4.1854,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363737525084138e-05,
+      "loss": 4.1914,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362898930333086e-05,
+      "loss": 4.1985,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620603355820335e-05,
+      "loss": 4.1824,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361221740830982e-05,
+      "loss": 4.1919,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36038314607993e-05,
+      "loss": 4.1953,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359544551328878e-05,
+      "loss": 4.1868,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358705956577826e-05,
+      "loss": 4.2157,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357867361826774e-05,
+      "loss": 4.1984,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357030404956095e-05,
+      "loss": 4.1943,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356191810205043e-05,
+      "loss": 4.1968,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355354853334364e-05,
+      "loss": 4.2077,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354516258583312e-05,
+      "loss": 4.1861,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35367766383226e-05,
+      "loss": 4.1953,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352839069081208e-05,
+      "loss": 4.193,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352002112210529e-05,
+      "loss": 4.1881,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511635174594776e-05,
+      "loss": 4.181,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503249227084256e-05,
+      "loss": 4.196,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494863279573736e-05,
+      "loss": 4.204,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486477332063216e-05,
+      "loss": 4.1944,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478091384552696e-05,
+      "loss": 4.198,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469705437042175e-05,
+      "loss": 4.1927,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461335868335385e-05,
+      "loss": 4.1912,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3452949920824865e-05,
+      "loss": 4.1943,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444580352118074e-05,
+      "loss": 4.1811,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436194404607554e-05,
+      "loss": 4.1834,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427808457097034e-05,
+      "loss": 4.1783,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419422509586514e-05,
+      "loss": 4.182,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411036562075993e-05,
+      "loss": 4.1885,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402650614565473e-05,
+      "loss": 4.1908,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339426466705496e-05,
+      "loss": 4.1819,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338587871954444e-05,
+      "loss": 4.1862,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337749277203392e-05,
+      "loss": 4.1914,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33691068245234e-05,
+      "loss": 4.1873,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336072087701287e-05,
+      "loss": 4.1897,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335233492950235e-05,
+      "loss": 4.1842,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334396536079557e-05,
+      "loss": 4.169,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333559579208878e-05,
+      "loss": 4.1936,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332720984457826e-05,
+      "loss": 4.1784,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331882389706774e-05,
+      "loss": 4.1888,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331043794955722e-05,
+      "loss": 4.1751,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330206838085043e-05,
+      "loss": 4.1867,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3293682433339914e-05,
+      "loss": 4.1663,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285296485829394e-05,
+      "loss": 4.1907,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3276910538318874e-05,
+      "loss": 4.1652,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326852459080835e-05,
+      "loss": 4.1774,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326013864329783e-05,
+      "loss": 4.1834,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325175269578731e-05,
+      "loss": 4.1716,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324336674827679e-05,
+      "loss": 4.1712,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323498080076627e-05,
+      "loss": 4.1831,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3226611232059476e-05,
+      "loss": 4.1617,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218225284548956e-05,
+      "loss": 4.1614,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209839337038436e-05,
+      "loss": 4.1763,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201453389527916e-05,
+      "loss": 4.1803,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193067442017396e-05,
+      "loss": 4.1648,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318469787331061e-05,
+      "loss": 4.1787,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317631192580009e-05,
+      "loss": 4.1543,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316792597828957e-05,
+      "loss": 4.1708,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315954003077905e-05,
+      "loss": 4.1787,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315115408326853e-05,
+      "loss": 4.1773,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314276813575801e-05,
+      "loss": 4.1805,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313438218824749e-05,
+      "loss": 4.1828,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312599624073697e-05,
+      "loss": 4.191,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311764305083391e-05,
+      "loss": 4.1606,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310925710332339e-05,
+      "loss": 4.1763,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310087115581287e-05,
+      "loss": 4.1795,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309248520830235e-05,
+      "loss": 4.1537,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3084115639595565e-05,
+      "loss": 4.178,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075729692085045e-05,
+      "loss": 4.1722,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067343744574525e-05,
+      "loss": 4.1771,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058957797064005e-05,
+      "loss": 4.1781,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050571849553485e-05,
+      "loss": 4.1627,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042202280846694e-05,
+      "loss": 4.1696,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033816333336174e-05,
+      "loss": 4.1715,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025430385825654e-05,
+      "loss": 4.1815,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3017044438315134e-05,
+      "loss": 4.1695,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008658490804614e-05,
+      "loss": 4.1805,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000272543294094e-05,
+      "loss": 4.1805,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991886595783574e-05,
+      "loss": 4.1647,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983500648273054e-05,
+      "loss": 4.1728,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297513107956627e-05,
+      "loss": 4.1632,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296674513205575e-05,
+      "loss": 4.1746,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295835918454523e-05,
+      "loss": 4.184,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294997323703471e-05,
+      "loss": 4.1608,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294162004713165e-05,
+      "loss": 4.1649,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293323409962113e-05,
+      "loss": 4.1597,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292484815211061e-05,
+      "loss": 4.173,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291646220460009e-05,
+      "loss": 4.1803,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290807625708957e-05,
+      "loss": 4.1646,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289969030957905e-05,
+      "loss": 4.16,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289130436206853e-05,
+      "loss": 4.1532,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288291841455801e-05,
+      "loss": 4.1759,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287454884585122e-05,
+      "loss": 4.1461,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28661628983407e-05,
+      "loss": 4.1645,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285779332963391e-05,
+      "loss": 4.1622,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284940738212339e-05,
+      "loss": 4.1654,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28410378134166e-05,
+      "loss": 4.161,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283265186590608e-05,
+      "loss": 4.1673,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282426591839556e-05,
+      "loss": 4.162,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281587997088504e-05,
+      "loss": 4.1559,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280749402337452e-05,
+      "loss": 4.1657,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799108075864e-05,
+      "loss": 4.165,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790722128353474e-05,
+      "loss": 4.1699,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278233618084296e-05,
+      "loss": 4.1716,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277396661213618e-05,
+      "loss": 4.163,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276558066462566e-05,
+      "loss": 4.1526,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275719471711513e-05,
+      "loss": 4.1609,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274880876960461e-05,
+      "loss": 4.1652,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274042282209409e-05,
+      "loss": 4.1518,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273203687458357e-05,
+      "loss": 4.1652,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272365092707305e-05,
+      "loss": 4.1507,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271526497956253e-05,
+      "loss": 4.1654,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270689541085574e-05,
+      "loss": 4.1558,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269850946334522e-05,
+      "loss": 4.1589,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26901235158347e-05,
+      "loss": 4.1463,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268173756832418e-05,
+      "loss": 4.157,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2673367999617395e-05,
+      "loss": 4.1589,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2664982052106875e-05,
+      "loss": 4.1638,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2656596104596355e-05,
+      "loss": 4.159,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2648210157085835e-05,
+      "loss": 4.1662,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639824209575315e-05,
+      "loss": 4.1483,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631454640868524e-05,
+      "loss": 4.16,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623068693358004e-05,
+      "loss": 4.1675,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2614682745847484e-05,
+      "loss": 4.1654,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606296798336964e-05,
+      "loss": 4.1632,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597910850826443e-05,
+      "loss": 4.1674,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589524903315923e-05,
+      "loss": 4.1557,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581138955805403e-05,
+      "loss": 4.1648,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257275300829488e-05,
+      "loss": 4.1454,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256436706078436e-05,
+      "loss": 4.1567,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255599749207758e-05,
+      "loss": 4.1511,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254762792337079e-05,
+      "loss": 4.153,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253924197586027e-05,
+      "loss": 4.1541,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253085602834975e-05,
+      "loss": 4.1486,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252247008083923e-05,
+      "loss": 4.1597,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251408413332871e-05,
+      "loss": 4.1447,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250569818581819e-05,
+      "loss": 4.1659,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.1535773277282715,
+      "eval_runtime": 302.8427,
+      "eval_samples_per_second": 1260.03,
+      "eval_steps_per_second": 39.377,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249731223830766e-05,
+      "loss": 4.1472,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248892629079714e-05,
+      "loss": 4.1437,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248054034328662e-05,
+      "loss": 4.1562,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24721543957761e-05,
+      "loss": 4.1526,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246376844826558e-05,
+      "loss": 4.1555,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245538250075506e-05,
+      "loss": 4.1467,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244699655324455e-05,
+      "loss": 4.1367,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243861060573403e-05,
+      "loss": 4.1481,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243022465822351e-05,
+      "loss": 4.1502,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242185508951672e-05,
+      "loss": 4.1578,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24134691420062e-05,
+      "loss": 4.1502,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240508319449568e-05,
+      "loss": 4.1543,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239669724698516e-05,
+      "loss": 4.1401,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388311299474637e-05,
+      "loss": 4.1429,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379925351964117e-05,
+      "loss": 4.1496,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371539404453596e-05,
+      "loss": 4.1348,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363153456943076e-05,
+      "loss": 4.1439,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235476750943255e-05,
+      "loss": 4.1492,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234638156192203e-05,
+      "loss": 4.1417,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337995614411516e-05,
+      "loss": 4.1656,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329609666900996e-05,
+      "loss": 4.1554,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321223719390476e-05,
+      "loss": 4.1436,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312837771879956e-05,
+      "loss": 4.1498,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304468203173165e-05,
+      "loss": 4.1629,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296082255662645e-05,
+      "loss": 4.1391,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2287696308152125e-05,
+      "loss": 4.1507,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279310360641605e-05,
+      "loss": 4.1426,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2270924413131085e-05,
+      "loss": 4.1423,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262538465620565e-05,
+      "loss": 4.1371,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254152518110045e-05,
+      "loss": 4.1478,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245766570599525e-05,
+      "loss": 4.1553,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237397001892734e-05,
+      "loss": 4.1466,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229011054382214e-05,
+      "loss": 4.1534,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22206251068717e-05,
+      "loss": 4.1496,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221225553816491e-05,
+      "loss": 4.1426,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220386959065439e-05,
+      "loss": 4.1475,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219548364314387e-05,
+      "loss": 4.1377,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218711407443708e-05,
+      "loss": 4.1358,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217872812692656e-05,
+      "loss": 4.1343,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217034217941604e-05,
+      "loss": 4.1343,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216195623190552e-05,
+      "loss": 4.1412,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2153570284395e-05,
+      "loss": 4.1468,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214518433688448e-05,
+      "loss": 4.1361,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213679838937396e-05,
+      "loss": 4.1372,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212841244186344e-05,
+      "loss": 4.148,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212002649435292e-05,
+      "loss": 4.1409,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21116405468424e-05,
+      "loss": 4.1465,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210325459933188e-05,
+      "loss": 4.1389,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209486865182136e-05,
+      "loss": 4.1201,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208649908311457e-05,
+      "loss": 4.1515,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207812951440778e-05,
+      "loss": 4.1314,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206974356689726e-05,
+      "loss": 4.1451,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061357619386736e-05,
+      "loss": 4.134,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052971671876216e-05,
+      "loss": 4.1357,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204460210316943e-05,
+      "loss": 4.1289,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203621615565891e-05,
+      "loss": 4.1424,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027830208148385e-05,
+      "loss": 4.1245,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201944426063787e-05,
+      "loss": 4.1331,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201105831312735e-05,
+      "loss": 4.1429,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200267236561683e-05,
+      "loss": 4.1234,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199428641810631e-05,
+      "loss": 4.1269,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198590047059579e-05,
+      "loss": 4.1443,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197751452308527e-05,
+      "loss": 4.1143,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196912857557475e-05,
+      "loss": 4.1184,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196074262806423e-05,
+      "loss": 4.1328,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195235668055371e-05,
+      "loss": 4.1364,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194398711184692e-05,
+      "loss": 4.1246,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193561754314013e-05,
+      "loss": 4.1364,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192723159562961e-05,
+      "loss": 4.1147,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191884564811909e-05,
+      "loss": 4.1259,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191045970060857e-05,
+      "loss": 4.1367,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902073753098057e-05,
+      "loss": 4.1336,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1893704184391266e-05,
+      "loss": 4.1376,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885318236880746e-05,
+      "loss": 4.1378,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876932289370226e-05,
+      "loss": 4.1461,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868562720663435e-05,
+      "loss": 4.1192,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860176773152915e-05,
+      "loss": 4.1307,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851790825642395e-05,
+      "loss": 4.1363,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843404878131875e-05,
+      "loss": 4.1131,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835018930621355e-05,
+      "loss": 4.1338,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826632983110834e-05,
+      "loss": 4.1305,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818247035600314e-05,
+      "loss": 4.1354,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809861088089794e-05,
+      "loss": 4.1347,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801475140579274e-05,
+      "loss": 4.1227,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793089193068754e-05,
+      "loss": 4.1257,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784703245558234e-05,
+      "loss": 4.1299,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776317298047714e-05,
+      "loss": 4.1333,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176794772934093e-05,
+      "loss": 4.131,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17595617818304e-05,
+      "loss": 4.138,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175119221312362e-05,
+      "loss": 4.1372,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17428062656131e-05,
+      "loss": 4.1251,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173442031810257e-05,
+      "loss": 4.13,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172603437059205e-05,
+      "loss": 4.1196,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171764842308153e-05,
+      "loss": 4.1384,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170926247557101e-05,
+      "loss": 4.1376,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170087652806049e-05,
+      "loss": 4.1232,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169249058054998e-05,
+      "loss": 4.1217,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168412101184319e-05,
+      "loss": 4.1168,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167573506433267e-05,
+      "loss": 4.1331,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166734911682215e-05,
+      "loss": 4.1385,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165897954811536e-05,
+      "loss": 4.1214,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165059360060484e-05,
+      "loss": 4.1216,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164220765309432e-05,
+      "loss": 4.1165,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16338217055838e-05,
+      "loss": 4.1286,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162543575807328e-05,
+      "loss": 4.1096,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161704981056276e-05,
+      "loss": 4.1227,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160866386305224e-05,
+      "loss": 4.1146,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160027791554172e-05,
+      "loss": 4.128,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1591908346834926e-05,
+      "loss": 4.1187,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158353877812814e-05,
+      "loss": 4.1262,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157515283061762e-05,
+      "loss": 4.1208,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15667668831071e-05,
+      "loss": 4.113,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155838093559658e-05,
+      "loss": 4.1269,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154999498808606e-05,
+      "loss": 4.1244,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154162541937927e-05,
+      "loss": 4.1295,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153323947186875e-05,
+      "loss": 4.1313,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152485352435823e-05,
+      "loss": 4.1185,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151646757684771e-05,
+      "loss": 4.1215,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150808162933719e-05,
+      "loss": 4.1144,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149969568182667e-05,
+      "loss": 4.1259,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149130973431615e-05,
+      "loss": 4.1186,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148292378680563e-05,
+      "loss": 4.1206,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474554218098846e-05,
+      "loss": 4.1073,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466184649392055e-05,
+      "loss": 4.1238,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457815080685264e-05,
+      "loss": 4.1138,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449429133174744e-05,
+      "loss": 4.1206,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1441043185664224e-05,
+      "loss": 4.1038,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432657238153704e-05,
+      "loss": 4.1206,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424271290643184e-05,
+      "loss": 4.1154,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415885343132664e-05,
+      "loss": 4.1294,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407499395622144e-05,
+      "loss": 4.1181,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399113448111624e-05,
+      "loss": 4.125,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390727500601104e-05,
+      "loss": 4.1083,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382341553090584e-05,
+      "loss": 4.1235,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1373955605580064e-05,
+      "loss": 4.1275,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1365569658069544e-05,
+      "loss": 4.1249,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135720008936276e-05,
+      "loss": 4.1238,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134881414185223e-05,
+      "loss": 4.1277,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134042819434171e-05,
+      "loss": 4.1157,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133204224683119e-05,
+      "loss": 4.1267,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132365629932067e-05,
+      "loss": 4.107,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131528673061388e-05,
+      "loss": 4.1181,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130690078310336e-05,
+      "loss": 4.1118,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129853121439658e-05,
+      "loss": 4.1145,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129014526688606e-05,
+      "loss": 4.1105,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128175931937554e-05,
+      "loss": 4.1088,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127337337186502e-05,
+      "loss": 4.1195,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126500380315823e-05,
+      "loss": 4.1104,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1256617855647707e-05,
+      "loss": 4.1289,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.122705459594727,
+      "eval_runtime": 295.5193,
+      "eval_samples_per_second": 1291.256,
+      "eval_steps_per_second": 40.353,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1248231908137186e-05,
+      "loss": 4.1089,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1239845960626666e-05,
+      "loss": 4.1035,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1231460013116146e-05,
+      "loss": 4.1183,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1223074065605626e-05,
+      "loss": 4.1133,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1214704496898835e-05,
+      "loss": 4.1119,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206318549388315e-05,
+      "loss": 4.113,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1197932601877795e-05,
+      "loss": 4.1042,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1189546654367275e-05,
+      "loss": 4.1071,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1181160706856755e-05,
+      "loss": 4.1134,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117277475934624e-05,
+      "loss": 4.1207,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116438881183572e-05,
+      "loss": 4.1122,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115601924312893e-05,
+      "loss": 4.1114,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114763329561841e-05,
+      "loss": 4.1078,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113924734810789e-05,
+      "loss": 4.1046,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113086140059737e-05,
+      "loss": 4.1086,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112247545308685e-05,
+      "loss": 4.0964,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111410588438006e-05,
+      "loss": 4.1082,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110571993686954e-05,
+      "loss": 4.1121,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109735036816275e-05,
+      "loss": 4.1023,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108896442065223e-05,
+      "loss": 4.1257,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108057847314171e-05,
+      "loss": 4.1214,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072192525631196e-05,
+      "loss": 4.1064,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063806578120676e-05,
+      "loss": 4.1121,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1055420630610156e-05,
+      "loss": 4.123,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1047034683099636e-05,
+      "loss": 4.1015,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1038648735589115e-05,
+      "loss": 4.1138,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030279166882325e-05,
+      "loss": 4.1069,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021893219371805e-05,
+      "loss": 4.1047,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1013507271861285e-05,
+      "loss": 4.0987,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005121324350764e-05,
+      "loss": 4.1137,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0996735376840244e-05,
+      "loss": 4.1167,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988365808133454e-05,
+      "loss": 4.1138,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0979979860622933e-05,
+      "loss": 4.116,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0971593913112413e-05,
+      "loss": 4.107,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096320796560189e-05,
+      "loss": 4.1053,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095483839689511e-05,
+      "loss": 4.1126,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094645244938459e-05,
+      "loss": 4.1032,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09380828806778e-05,
+      "loss": 4.0947,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092969693316728e-05,
+      "loss": 4.0977,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092131098565676e-05,
+      "loss": 4.0984,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091292503814624e-05,
+      "loss": 4.1027,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090453909063572e-05,
+      "loss": 4.1103,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089615314312519e-05,
+      "loss": 4.1032,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088776719561467e-05,
+      "loss": 4.0982,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087938124810415e-05,
+      "loss": 4.1121,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087101167939737e-05,
+      "loss": 4.1076,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086262573188685e-05,
+      "loss": 4.108,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085423978437633e-05,
+      "loss": 4.1047,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084585383686581e-05,
+      "loss": 4.0793,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083746788935529e-05,
+      "loss": 4.1193,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082908194184477e-05,
+      "loss": 4.0957,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082069599433425e-05,
+      "loss": 4.1106,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081231004682373e-05,
+      "loss": 4.0963,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080392409931321e-05,
+      "loss": 4.1006,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0795554530606416e-05,
+      "loss": 4.0924,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787168583095896e-05,
+      "loss": 4.104,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0778799014389105e-05,
+      "loss": 4.0917,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770413066878585e-05,
+      "loss": 4.0959,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762027119368065e-05,
+      "loss": 4.109,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075364117185755e-05,
+      "loss": 4.0896,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074525522434703e-05,
+      "loss": 4.0877,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073686927683651e-05,
+      "loss": 4.1073,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072848332932599e-05,
+      "loss": 4.083,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072009738181547e-05,
+      "loss": 4.0815,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071172781310868e-05,
+      "loss": 4.0983,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070334186559816e-05,
+      "loss": 4.0997,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069495591808764e-05,
+      "loss": 4.0908,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068656997057712e-05,
+      "loss": 4.1007,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067820040187033e-05,
+      "loss": 4.0773,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066981445435981e-05,
+      "loss": 4.0919,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066142850684929e-05,
+      "loss": 4.1013,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065304255933877e-05,
+      "loss": 4.0959,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0644672990631985e-05,
+      "loss": 4.1058,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636287043121465e-05,
+      "loss": 4.1024,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627901095610945e-05,
+      "loss": 4.1085,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619531526904154e-05,
+      "loss": 4.0834,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611145579393634e-05,
+      "loss": 4.0989,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602759631883114e-05,
+      "loss": 4.0995,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594373684372594e-05,
+      "loss": 4.0815,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0585987736862074e-05,
+      "loss": 4.0977,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057761816815528e-05,
+      "loss": 4.0974,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056923222064476e-05,
+      "loss": 4.1008,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056084627313424e-05,
+      "loss": 4.1044,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055246032562372e-05,
+      "loss": 4.0833,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054409075691694e-05,
+      "loss": 4.0905,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053570480940642e-05,
+      "loss": 4.0961,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052733524069963e-05,
+      "loss": 4.1003,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051894929318911e-05,
+      "loss": 4.0991,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051056334567859e-05,
+      "loss": 4.0999,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050217739816807e-05,
+      "loss": 4.1023,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049379145065755e-05,
+      "loss": 4.0947,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048540550314702e-05,
+      "loss": 4.0946,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04770195556365e-05,
+      "loss": 4.0884,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046863360812598e-05,
+      "loss": 4.1015,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046024766061546e-05,
+      "loss": 4.1013,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045186171310494e-05,
+      "loss": 4.0929,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443492144398157e-05,
+      "loss": 4.0851,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435106196887637e-05,
+      "loss": 4.0837,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0426720249377116e-05,
+      "loss": 4.0975,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418350680670326e-05,
+      "loss": 4.1071,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0409964733159806e-05,
+      "loss": 4.0887,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0401578785649285e-05,
+      "loss": 4.0883,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0393192838138765e-05,
+      "loss": 4.0869,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0384806890628245e-05,
+      "loss": 4.0912,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0376420943117725e-05,
+      "loss": 4.0786,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0368034995607205e-05,
+      "loss": 4.0916,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0359649048096685e-05,
+      "loss": 4.0786,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351263100586165e-05,
+      "loss": 4.0926,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342877153075645e-05,
+      "loss": 4.0895,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0334491205565125e-05,
+      "loss": 4.0902,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326105258054605e-05,
+      "loss": 4.0863,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031773568934782e-05,
+      "loss": 4.086,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030936612064103e-05,
+      "loss": 4.0903,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030098017313051e-05,
+      "loss": 4.0914,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029259422561999e-05,
+      "loss": 4.0953,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028420827810947e-05,
+      "loss": 4.1023,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027582233059895e-05,
+      "loss": 4.0864,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026743638308843e-05,
+      "loss": 4.0893,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025905043557791e-05,
+      "loss": 4.083,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025068086687112e-05,
+      "loss": 4.0898,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02422949193606e-05,
+      "loss": 4.0879,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023390897185008e-05,
+      "loss": 4.0874,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022552302433956e-05,
+      "loss": 4.0759,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021713707682904e-05,
+      "loss": 4.0885,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0208767508122255e-05,
+      "loss": 4.0865,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200381560611735e-05,
+      "loss": 4.0859,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0191995613101214e-05,
+      "loss": 4.0679,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018360966559069e-05,
+      "loss": 4.0934,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017522371808017e-05,
+      "loss": 4.0811,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016683777056965e-05,
+      "loss": 4.0953,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015846820186286e-05,
+      "loss": 4.0879,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015008225435234e-05,
+      "loss": 4.0923,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0141696306841817e-05,
+      "loss": 4.0757,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0133310359331297e-05,
+      "loss": 4.0933,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012492441182078e-05,
+      "loss": 4.095,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011653846431026e-05,
+      "loss": 4.0927,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010815251679974e-05,
+      "loss": 4.0929,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009976656928922e-05,
+      "loss": 4.0931,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009139700058243e-05,
+      "loss": 4.0844,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008301105307191e-05,
+      "loss": 4.0942,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007462510556139e-05,
+      "loss": 4.0774,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006623915805087e-05,
+      "loss": 4.089,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005786958934408e-05,
+      "loss": 4.0804,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004948364183356e-05,
+      "loss": 4.0884,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004111407312677e-05,
+      "loss": 4.0747,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003272812561625e-05,
+      "loss": 4.0786,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002434217810574e-05,
+      "loss": 4.0837,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001595623059522e-05,
+      "loss": 4.0832,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00075702830847e-05,
+      "loss": 4.0969,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.099428653717041,
+      "eval_runtime": 294.2439,
+      "eval_samples_per_second": 1296.853,
+      "eval_steps_per_second": 40.528,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999918433557418e-05,
+      "loss": 4.0751,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999079838806366e-05,
+      "loss": 4.0768,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998241244055314e-05,
+      "loss": 4.08,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997402649304262e-05,
+      "loss": 4.0823,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99656405455321e-05,
+      "loss": 4.0854,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957254598021577e-05,
+      "loss": 4.078,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994886865051105e-05,
+      "loss": 4.0829,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994048270300053e-05,
+      "loss": 4.0761,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993209675549001e-05,
+      "loss": 4.0778,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992371080797949e-05,
+      "loss": 4.0932,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991532486046897e-05,
+      "loss": 4.0784,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990693891295845e-05,
+      "loss": 4.0814,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989855296544793e-05,
+      "loss": 4.0765,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890167017937416e-05,
+      "loss": 4.0699,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9881781070426896e-05,
+      "loss": 4.0824,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873395122916376e-05,
+      "loss": 4.0667,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865009175405856e-05,
+      "loss": 4.0787,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9856623227895336e-05,
+      "loss": 4.0826,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848237280384816e-05,
+      "loss": 4.0702,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983985133287429e-05,
+      "loss": 4.0927,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9831481764167505e-05,
+      "loss": 4.0935,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823095816656985e-05,
+      "loss": 4.0788,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981470986914646e-05,
+      "loss": 4.0788,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980632392163594e-05,
+      "loss": 4.0911,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9797954352929154e-05,
+      "loss": 4.0777,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9789568405418634e-05,
+      "loss": 4.0773,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781182457908114e-05,
+      "loss": 4.077,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9772796510397594e-05,
+      "loss": 4.0747,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976442694169081e-05,
+      "loss": 4.0762,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975604099418029e-05,
+      "loss": 4.0779,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974765504666976e-05,
+      "loss": 4.082,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973928547796298e-05,
+      "loss": 4.0857,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973089953045246e-05,
+      "loss": 4.0862,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972251358294193e-05,
+      "loss": 4.079,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971412763543141e-05,
+      "loss": 4.0741,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970574168792089e-05,
+      "loss": 4.0844,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969737211921411e-05,
+      "loss": 4.0747,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968898617170359e-05,
+      "loss": 4.066,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968060022419307e-05,
+      "loss": 4.0697,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967221427668255e-05,
+      "loss": 4.0669,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966382832917203e-05,
+      "loss": 4.0732,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965544238166151e-05,
+      "loss": 4.0785,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964705643415099e-05,
+      "loss": 4.0764,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963867048664047e-05,
+      "loss": 4.0735,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963028453912995e-05,
+      "loss": 4.0772,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9621914970423156e-05,
+      "loss": 4.0789,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613545401716366e-05,
+      "loss": 4.0744,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605159454205845e-05,
+      "loss": 4.0774,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9596773506695325e-05,
+      "loss": 4.0509,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9588387559184805e-05,
+      "loss": 4.0891,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958000161167429e-05,
+      "loss": 4.0667,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957161566416377e-05,
+      "loss": 4.084,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956322971665325e-05,
+      "loss": 4.0655,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955484376914273e-05,
+      "loss": 4.0693,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954647420043594e-05,
+      "loss": 4.0665,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953808825292542e-05,
+      "loss": 4.0733,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95297023054149e-05,
+      "loss": 4.0644,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952131635790438e-05,
+      "loss": 4.0671,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951294678919759e-05,
+      "loss": 4.0782,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95045772204908e-05,
+      "loss": 4.0649,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949619127298028e-05,
+      "loss": 4.0563,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948780532546976e-05,
+      "loss": 4.0807,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479419377959246e-05,
+      "loss": 4.054,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471033430448726e-05,
+      "loss": 4.0511,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462647482938206e-05,
+      "loss": 4.0704,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454261535427686e-05,
+      "loss": 4.0723,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445875587917166e-05,
+      "loss": 4.0648,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437506019210375e-05,
+      "loss": 4.069,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429120071699855e-05,
+      "loss": 4.0509,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420734124189335e-05,
+      "loss": 4.0615,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9412348176678815e-05,
+      "loss": 4.0736,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9403962229168295e-05,
+      "loss": 4.0669,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9395592660461504e-05,
+      "loss": 4.0789,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387206712950984e-05,
+      "loss": 4.0693,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9378820765440464e-05,
+      "loss": 4.0823,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937045119673368e-05,
+      "loss": 4.0558,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936206524922316e-05,
+      "loss": 4.0705,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935367930171264e-05,
+      "loss": 4.067,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934529335420212e-05,
+      "loss": 4.0583,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933690740669159e-05,
+      "loss": 4.068,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932852145918107e-05,
+      "loss": 4.0717,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932013551167055e-05,
+      "loss": 4.0699,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931176594296377e-05,
+      "loss": 4.0776,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930337999545324e-05,
+      "loss": 4.0536,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929499404794272e-05,
+      "loss": 4.0591,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92866081004322e-05,
+      "loss": 4.0718,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927822215292168e-05,
+      "loss": 4.0697,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926983620541116e-05,
+      "loss": 4.0707,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926145025790065e-05,
+      "loss": 4.0739,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925308068919386e-05,
+      "loss": 4.0695,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924469474168334e-05,
+      "loss": 4.0691,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923630879417282e-05,
+      "loss": 4.0658,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92279228466623e-05,
+      "loss": 4.0592,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9219553277955506e-05,
+      "loss": 4.0773,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211167330444986e-05,
+      "loss": 4.0694,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9202781382934466e-05,
+      "loss": 4.0668,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9194395435423946e-05,
+      "loss": 4.0614,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9186009487913426e-05,
+      "loss": 4.0539,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9177639919206635e-05,
+      "loss": 4.0673,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169253971696115e-05,
+      "loss": 4.0769,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91608680241856e-05,
+      "loss": 4.0621,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915248207667508e-05,
+      "loss": 4.0615,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914409612916456e-05,
+      "loss": 4.0589,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913572656045777e-05,
+      "loss": 4.0606,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912734061294725e-05,
+      "loss": 4.0582,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911895466543673e-05,
+      "loss": 4.0575,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911056871792621e-05,
+      "loss": 4.0505,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910218277041569e-05,
+      "loss": 4.068,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90938132017089e-05,
+      "loss": 4.0642,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908542725419838e-05,
+      "loss": 4.0623,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907704130668786e-05,
+      "loss": 4.0539,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906865535917734e-05,
+      "loss": 4.0629,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906026941166682e-05,
+      "loss": 4.0636,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90518834641563e-05,
+      "loss": 4.063,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9043497516645786e-05,
+      "loss": 4.0687,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035127947938995e-05,
+      "loss": 4.0728,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026758379232204e-05,
+      "loss": 4.0574,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901838881052541e-05,
+      "loss": 4.0655,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901000286301489e-05,
+      "loss": 4.0501,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900161691550437e-05,
+      "loss": 4.0635,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899323096799385e-05,
+      "loss": 4.0614,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898484502048333e-05,
+      "loss": 4.0581,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897645907297281e-05,
+      "loss": 4.0532,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896807312546229e-05,
+      "loss": 4.0619,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895968717795177e-05,
+      "loss": 4.0612,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895131760924499e-05,
+      "loss": 4.0583,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894293166173447e-05,
+      "loss": 4.0433,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893454571422395e-05,
+      "loss": 4.0626,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892615976671343e-05,
+      "loss": 4.0543,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89177738192029e-05,
+      "loss": 4.0678,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890938787169238e-05,
+      "loss": 4.0629,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890100192418186e-05,
+      "loss": 4.0607,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889261597667134e-05,
+      "loss": 4.0514,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888424640796455e-05,
+      "loss": 4.0671,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887587683925777e-05,
+      "loss": 4.0676,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886749089174725e-05,
+      "loss": 4.0668,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885910494423673e-05,
+      "loss": 4.0672,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885071899672621e-05,
+      "loss": 4.0655,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884233304921569e-05,
+      "loss": 4.0594,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8833947101705167e-05,
+      "loss": 4.066,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8825561154194647e-05,
+      "loss": 4.0494,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817175206684127e-05,
+      "loss": 4.0617,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8808805637977336e-05,
+      "loss": 4.0554,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8800419690466816e-05,
+      "loss": 4.0581,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792033742956296e-05,
+      "loss": 4.0489,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8783664174249505e-05,
+      "loss": 4.0517,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8775278226738985e-05,
+      "loss": 4.0579,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87669086580322e-05,
+      "loss": 4.0589,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875852271052168e-05,
+      "loss": 4.0686,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.08057165145874,
+      "eval_runtime": 294.6161,
+      "eval_samples_per_second": 1295.214,
+      "eval_steps_per_second": 40.476,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875013676301116e-05,
+      "loss": 4.045,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874175081550064e-05,
+      "loss": 4.0497,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873336486799012e-05,
+      "loss": 4.051,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.87249789204796e-05,
+      "loss": 4.0617,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871659297296908e-05,
+      "loss": 4.0584,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870820702545856e-05,
+      "loss": 4.0494,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869982107794804e-05,
+      "loss": 4.0545,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869143513043752e-05,
+      "loss": 4.0485,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683049182927e-05,
+      "loss": 4.0507,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867467961422021e-05,
+      "loss": 4.0681,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866629366670969e-05,
+      "loss": 4.0518,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865790771919917e-05,
+      "loss": 4.052,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864952177168865e-05,
+      "loss": 4.062,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864113582417813e-05,
+      "loss": 4.0368,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863274987666761e-05,
+      "loss": 4.0539,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862436392915709e-05,
+      "loss": 4.041,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861597798164657e-05,
+      "loss": 4.0533,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8607608412939785e-05,
+      "loss": 4.0552,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8599222465429265e-05,
+      "loss": 4.0441,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859083651791874e-05,
+      "loss": 4.0656,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858245057040822e-05,
+      "loss": 4.0673,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85740646228977e-05,
+      "loss": 4.0559,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8565695054190914e-05,
+      "loss": 4.0572,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855730910668039e-05,
+      "loss": 4.0593,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854892315916987e-05,
+      "loss": 4.0548,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854053721165935e-05,
+      "loss": 4.0494,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853216764295256e-05,
+      "loss": 4.0556,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852378169544204e-05,
+      "loss": 4.0504,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851539574793152e-05,
+      "loss": 4.0489,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8507009800421e-05,
+      "loss": 4.0523,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849864023171421e-05,
+      "loss": 4.0546,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849027066300743e-05,
+      "loss": 4.0636,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8481901094300636e-05,
+      "loss": 4.0577,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8473515146790116e-05,
+      "loss": 4.0564,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8465129199279596e-05,
+      "loss": 4.0484,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8456743251769076e-05,
+      "loss": 4.0514,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448357304258556e-05,
+      "loss": 4.0535,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8439971356748036e-05,
+      "loss": 4.0431,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8431585409237516e-05,
+      "loss": 4.0421,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423199461726996e-05,
+      "loss": 4.0397,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414813514216476e-05,
+      "loss": 4.0484,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8406427566705956e-05,
+      "loss": 4.0497,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8398041619195436e-05,
+      "loss": 4.0497,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8389655671684916e-05,
+      "loss": 4.0491,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8381269724174396e-05,
+      "loss": 4.0543,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8372900155467605e-05,
+      "loss": 4.0499,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8364514207957085e-05,
+      "loss": 4.0492,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8356144639250294e-05,
+      "loss": 4.056,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8347758691739774e-05,
+      "loss": 4.0243,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8339372744229254e-05,
+      "loss": 4.0566,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833098679671874e-05,
+      "loss": 4.0438,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832260084920822e-05,
+      "loss": 4.0602,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83142149016977e-05,
+      "loss": 4.0423,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830582895418718e-05,
+      "loss": 4.0445,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829744300667666e-05,
+      "loss": 4.0365,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828905705916614e-05,
+      "loss": 4.0471,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828068749045935e-05,
+      "loss": 4.0442,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827230154294883e-05,
+      "loss": 4.0383,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826391559543831e-05,
+      "loss": 4.0531,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825552964792779e-05,
+      "loss": 4.043,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824714370041727e-05,
+      "loss": 4.0249,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823877413171048e-05,
+      "loss": 4.0556,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230404563003694e-05,
+      "loss": 4.0331,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8222018615493174e-05,
+      "loss": 4.0255,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8213632667982654e-05,
+      "loss": 4.0448,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8205246720472134e-05,
+      "loss": 4.045,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8196860772961614e-05,
+      "loss": 4.034,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8188474825451094e-05,
+      "loss": 4.0501,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8180088877940574e-05,
+      "loss": 4.0276,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817170293043005e-05,
+      "loss": 4.0347,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816333336172326e-05,
+      "loss": 4.0488,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815494741421274e-05,
+      "loss": 4.0408,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814657784550595e-05,
+      "loss": 4.0564,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813819189799543e-05,
+      "loss": 4.0435,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812980595048491e-05,
+      "loss": 4.0542,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812143638177813e-05,
+      "loss": 4.0352,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811305043426761e-05,
+      "loss": 4.0491,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810466448675709e-05,
+      "loss": 4.0356,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809627853924657e-05,
+      "loss": 4.0376,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808789259173605e-05,
+      "loss": 4.0402,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807950664422552e-05,
+      "loss": 4.0441,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071120696715e-05,
+      "loss": 4.0458,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806273474920448e-05,
+      "loss": 4.0554,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805436518049769e-05,
+      "loss": 4.0283,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804597923298717e-05,
+      "loss": 4.0351,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803759328547665e-05,
+      "loss": 4.0474,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802920733796613e-05,
+      "loss": 4.0466,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802082139045561e-05,
+      "loss": 4.0429,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8012435442945097e-05,
+      "loss": 4.0522,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8004065874238306e-05,
+      "loss": 4.0436,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7995679926727786e-05,
+      "loss": 4.0451,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7987293979217266e-05,
+      "loss": 4.0422,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7978908031706746e-05,
+      "loss": 4.0344,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7970522084196225e-05,
+      "loss": 4.0478,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7962136136685705e-05,
+      "loss": 4.0478,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7953750189175185e-05,
+      "loss": 4.0408,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7945364241664665e-05,
+      "loss": 4.039,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7936994672957874e-05,
+      "loss": 4.0268,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7928608725447354e-05,
+      "loss": 4.0431,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7920239156740564e-05,
+      "loss": 4.0529,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791185320923005e-05,
+      "loss": 4.0389,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790348364052326e-05,
+      "loss": 4.0385,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789509769301274e-05,
+      "loss": 4.0332,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788671174550222e-05,
+      "loss": 4.0324,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78783257979917e-05,
+      "loss": 4.0375,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786993985048118e-05,
+      "loss": 4.0317,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786155390297066e-05,
+      "loss": 4.0285,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785316795546014e-05,
+      "loss": 4.0429,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784478200794962e-05,
+      "loss": 4.0435,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78363960604391e-05,
+      "loss": 4.0328,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782801011292858e-05,
+      "loss": 4.033,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781964054422179e-05,
+      "loss": 4.035,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781125459671127e-05,
+      "loss": 4.0436,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780286864920075e-05,
+      "loss": 4.0361,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7794482701690235e-05,
+      "loss": 4.0473,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778609675417971e-05,
+      "loss": 4.0472,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7777727185472924e-05,
+      "loss": 4.034,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7769341237962404e-05,
+      "loss": 4.0416,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776097166925561e-05,
+      "loss": 4.0302,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775258572174509e-05,
+      "loss": 4.039,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774419977423457e-05,
+      "loss": 4.0419,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773581382672405e-05,
+      "loss": 4.0317,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7727427879213526e-05,
+      "loss": 4.0269,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7719041931703006e-05,
+      "loss": 4.0393,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771067236299622e-05,
+      "loss": 4.0375,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77022864154857e-05,
+      "loss": 4.037,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769390046797518e-05,
+      "loss": 4.0192,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768551452046466e-05,
+      "loss": 4.0356,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767712857295414e-05,
+      "loss": 4.0299,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766874262544362e-05,
+      "loss": 4.0449,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76603566779331e-05,
+      "loss": 4.041,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765197073042258e-05,
+      "loss": 4.0372,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764360116171579e-05,
+      "loss": 4.0249,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763521521420527e-05,
+      "loss": 4.0472,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762682926669475e-05,
+      "loss": 4.0396,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761844331918423e-05,
+      "loss": 4.0444,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761005737167371e-05,
+      "loss": 4.0394,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760167142416319e-05,
+      "loss": 4.0409,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7593301855456406e-05,
+      "loss": 4.0362,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7584915907945886e-05,
+      "loss": 4.0483,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7576529960435366e-05,
+      "loss": 4.0226,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7568144012924846e-05,
+      "loss": 4.041,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7559774444218055e-05,
+      "loss": 4.0311,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7551388496707535e-05,
+      "loss": 4.0339,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7543002549197015e-05,
+      "loss": 4.0284,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7534616601686495e-05,
+      "loss": 4.0287,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7526230654175975e-05,
+      "loss": 4.0335,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7517844706665455e-05,
+      "loss": 4.0368,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7509458759154935e-05,
+      "loss": 4.042,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.06446647644043,
+      "eval_runtime": 295.0472,
+      "eval_samples_per_second": 1293.322,
+      "eval_steps_per_second": 40.417,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7501072811644415e-05,
+      "loss": 4.0223,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749268686413389e-05,
+      "loss": 4.0293,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7484300916623375e-05,
+      "loss": 4.0268,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7475914969112855e-05,
+      "loss": 4.0393,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467529021602335e-05,
+      "loss": 4.0396,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459143074091815e-05,
+      "loss": 4.0265,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7450757126581294e-05,
+      "loss": 4.0305,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7442371179070774e-05,
+      "loss": 4.0261,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7433985231560254e-05,
+      "loss": 4.0311,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7425599284049734e-05,
+      "loss": 4.0443,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417213336539214e-05,
+      "loss": 4.0284,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7408827389028694e-05,
+      "loss": 4.0329,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7400441441518174e-05,
+      "loss": 4.0386,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7392055494007654e-05,
+      "loss": 4.0182,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738366954649713e-05,
+      "loss": 4.0289,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737528359898661e-05,
+      "loss": 4.0177,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736689765147609e-05,
+      "loss": 4.0306,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7358511703965574e-05,
+      "loss": 4.0343,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7350125756455054e-05,
+      "loss": 4.0204,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734175618774826e-05,
+      "loss": 4.0405,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733337024023774e-05,
+      "loss": 4.0443,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732498429272722e-05,
+      "loss": 4.0327,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73165983452167e-05,
+      "loss": 4.0369,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730821239770618e-05,
+      "loss": 4.0365,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729982645019566e-05,
+      "loss": 4.0358,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729144050268514e-05,
+      "loss": 4.028,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728305455517462e-05,
+      "loss": 4.0293,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727468498646783e-05,
+      "loss": 4.0335,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726631541776104e-05,
+      "loss": 4.0246,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725792947025053e-05,
+      "loss": 4.0288,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724954352274001e-05,
+      "loss": 4.034,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724115757522949e-05,
+      "loss": 4.0338,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723277162771897e-05,
+      "loss": 4.0427,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722440205901218e-05,
+      "loss": 4.0319,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7216016111501657e-05,
+      "loss": 4.0266,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7207630163991137e-05,
+      "loss": 4.0304,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7199260595284346e-05,
+      "loss": 4.0337,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7190874647773826e-05,
+      "loss": 4.0189,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7182488700263306e-05,
+      "loss": 4.0194,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7174102752752785e-05,
+      "loss": 4.0239,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7165716805242265e-05,
+      "loss": 4.0236,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157330857731745e-05,
+      "loss": 4.0312,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7148944910221225e-05,
+      "loss": 4.0258,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140558962710705e-05,
+      "loss": 4.0306,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713220577280765e-05,
+      "loss": 4.026,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712381982529713e-05,
+      "loss": 4.0337,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711543387778661e-05,
+      "loss": 4.0255,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710704793027609e-05,
+      "loss": 4.0358,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709866198276557e-05,
+      "loss": 4.0041,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709027603525505e-05,
+      "loss": 4.0325,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708189008774453e-05,
+      "loss": 4.0231,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707350414023401e-05,
+      "loss": 4.0387,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706511819272349e-05,
+      "loss": 4.0215,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705673224521296e-05,
+      "loss": 4.0212,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704834629770244e-05,
+      "loss": 4.02,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703996035019193e-05,
+      "loss": 4.025,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7031590781485146e-05,
+      "loss": 4.0301,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702320483397462e-05,
+      "loss": 4.0113,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70148188864641e-05,
+      "loss": 4.0359,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700643293895358e-05,
+      "loss": 4.0219,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699804699144306e-05,
+      "loss": 4.0018,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698967742273627e-05,
+      "loss": 4.035,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698129147522575e-05,
+      "loss": 4.0141,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6972921906518964e-05,
+      "loss": 4.0048,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696453595900844e-05,
+      "loss": 4.0274,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695615001149792e-05,
+      "loss": 4.0228,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69477640639874e-05,
+      "loss": 4.0111,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6939378116476884e-05,
+      "loss": 4.0318,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6930992168966363e-05,
+      "loss": 4.0035,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922606221455843e-05,
+      "loss": 4.0131,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691422027394532e-05,
+      "loss": 4.0263,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690585070523853e-05,
+      "loss": 4.0246,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689746475772801e-05,
+      "loss": 4.0315,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688907881021749e-05,
+      "loss": 4.023,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688069286270697e-05,
+      "loss": 4.0342,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687232329400018e-05,
+      "loss": 4.014,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686393734648966e-05,
+      "loss": 4.0299,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685555139897914e-05,
+      "loss": 4.0134,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684716545146862e-05,
+      "loss": 4.018,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68387795039581e-05,
+      "loss": 4.0216,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683040993525132e-05,
+      "loss": 4.0203,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68220239877408e-05,
+      "loss": 4.0296,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681363804023028e-05,
+      "loss": 4.0327,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6805268471523486e-05,
+      "loss": 4.015,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6796882524012966e-05,
+      "loss": 4.0136,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6788496576502446e-05,
+      "loss": 4.0222,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6780110628991926e-05,
+      "loss": 4.0286,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6771724681481406e-05,
+      "loss": 4.0203,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6763338733970886e-05,
+      "loss": 4.0312,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6754969165264095e-05,
+      "loss": 4.0254,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6746583217753575e-05,
+      "loss": 4.0258,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738197270243055e-05,
+      "loss": 4.0192,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729811322732535e-05,
+      "loss": 4.0145,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672142537522202e-05,
+      "loss": 4.0275,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67130394277115e-05,
+      "loss": 4.0246,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670465348020098e-05,
+      "loss": 4.0235,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696267532690455e-05,
+      "loss": 4.0178,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66879143427874e-05,
+      "loss": 4.0065,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667952839527688e-05,
+      "loss": 4.0213,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667114244776636e-05,
+      "loss": 4.0322,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666275650025584e-05,
+      "loss": 4.0198,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665437055274532e-05,
+      "loss": 4.0184,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66459846052348e-05,
+      "loss": 4.0164,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663759865772427e-05,
+      "loss": 4.0113,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662921271021376e-05,
+      "loss": 4.021,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6620843141506975e-05,
+      "loss": 4.0088,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6612457193996455e-05,
+      "loss": 4.0127,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6604087625289664e-05,
+      "loss": 4.0201,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6595701677779144e-05,
+      "loss": 4.0224,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6587315730268624e-05,
+      "loss": 4.017,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65789297827581e-05,
+      "loss": 4.0107,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657054383524758e-05,
+      "loss": 4.0132,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656215788773706e-05,
+      "loss": 4.0239,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655377194022654e-05,
+      "loss": 4.0161,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654538599271602e-05,
+      "loss": 4.0318,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6537016424009226e-05,
+      "loss": 4.0237,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652864685530244e-05,
+      "loss": 4.0175,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652026090779192e-05,
+      "loss": 4.0173,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65118749602814e-05,
+      "loss": 4.0117,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650350539157462e-05,
+      "loss": 4.0197,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64951194440641e-05,
+      "loss": 4.0253,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648673349655357e-05,
+      "loss": 4.0094,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647834754904305e-05,
+      "loss": 4.0083,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646996160153253e-05,
+      "loss": 4.018,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646157565402201e-05,
+      "loss": 4.0185,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645318970651149e-05,
+      "loss": 4.0142,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644480375900097e-05,
+      "loss": 4.0039,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643641781149045e-05,
+      "loss": 4.017,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642803186397993e-05,
+      "loss": 4.0065,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641964591646941e-05,
+      "loss": 4.029,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641125996895889e-05,
+      "loss": 4.0184,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6402890400252107e-05,
+      "loss": 4.0189,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6394520831545316e-05,
+      "loss": 4.0069,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6386134884034796e-05,
+      "loss": 4.0275,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6377748936524276e-05,
+      "loss": 4.0179,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6369362989013756e-05,
+      "loss": 4.0265,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6360977041503236e-05,
+      "loss": 4.0267,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6352607472796445e-05,
+      "loss": 4.0205,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6344221525285925e-05,
+      "loss": 4.0154,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6335835577775405e-05,
+      "loss": 4.03,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6327449630264884e-05,
+      "loss": 4.004,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6319063682754364e-05,
+      "loss": 4.0202,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310677735243844e-05,
+      "loss": 4.0121,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630229178773333e-05,
+      "loss": 4.0152,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629390584022281e-05,
+      "loss": 4.0088,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628553627151602e-05,
+      "loss": 4.0087,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.62771503240055e-05,
+      "loss": 4.0098,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626876437649498e-05,
+      "loss": 4.0218,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626037842898446e-05,
+      "loss": 4.0214,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.052252769470215,
+      "eval_runtime": 295.0368,
+      "eval_samples_per_second": 1293.368,
+      "eval_steps_per_second": 40.419,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625199248147393e-05,
+      "loss": 4.0101,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624360653396341e-05,
+      "loss": 4.0063,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623522058645289e-05,
+      "loss": 4.0061,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622683463894237e-05,
+      "loss": 4.0213,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621844869143185e-05,
+      "loss": 4.018,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62100955015288e-05,
+      "loss": 4.0095,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6201709554018285e-05,
+      "loss": 4.0121,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619332360650776e-05,
+      "loss": 4.0093,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618493765899724e-05,
+      "loss": 4.0094,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6176568090290454e-05,
+      "loss": 4.0192,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6168182142779934e-05,
+      "loss": 4.0152,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615979619526941e-05,
+      "loss": 4.0109,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615141024775889e-05,
+      "loss": 4.0205,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614302430024837e-05,
+      "loss": 3.9956,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613463835273785e-05,
+      "loss": 4.0113,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612625240522733e-05,
+      "loss": 3.9945,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611786645771681e-05,
+      "loss": 4.0159,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610948051020629e-05,
+      "loss": 4.0185,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610109456269577e-05,
+      "loss": 3.9966,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609272499398898e-05,
+      "loss": 4.0196,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608433904647846e-05,
+      "loss": 4.0252,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607595309896794e-05,
+      "loss": 4.0134,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606756715145742e-05,
+      "loss": 4.019,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605919758275063e-05,
+      "loss": 4.0177,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605081163524011e-05,
+      "loss": 4.0158,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604242568772959e-05,
+      "loss": 4.0105,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603403974021907e-05,
+      "loss": 4.0059,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602565379270855e-05,
+      "loss": 4.014,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601726784519803e-05,
+      "loss": 4.0076,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600888189768751e-05,
+      "loss": 4.0113,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600049595017699e-05,
+      "loss": 4.0146,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599212638147021e-05,
+      "loss": 4.0208,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5983756812763416e-05,
+      "loss": 4.0139,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5975370865252896e-05,
+      "loss": 4.0186,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5966984917742376e-05,
+      "loss": 4.004,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5958598970231856e-05,
+      "loss": 4.0133,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5950229401525065e-05,
+      "loss": 4.0168,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5941843454014545e-05,
+      "loss": 3.9954,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5933457506504025e-05,
+      "loss": 4.0046,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5925087937797234e-05,
+      "loss": 4.0056,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5916701990286714e-05,
+      "loss": 4.0042,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5908316042776194e-05,
+      "loss": 4.0105,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5899930095265674e-05,
+      "loss": 4.0096,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5891544147755154e-05,
+      "loss": 4.0088,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588315820024464e-05,
+      "loss": 4.0114,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587477225273412e-05,
+      "loss": 4.0128,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5866386305223594e-05,
+      "loss": 4.0062,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5858000357713074e-05,
+      "loss": 4.0164,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5849614410202554e-05,
+      "loss": 3.9906,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5841228462692034e-05,
+      "loss": 4.013,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5832842515181514e-05,
+      "loss": 4.0019,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582447294647472e-05,
+      "loss": 4.0175,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58160869989642e-05,
+      "loss": 4.005,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580770105145368e-05,
+      "loss": 3.9985,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579931510394316e-05,
+      "loss": 4.003,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579094553523638e-05,
+      "loss": 4.006,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578255958772586e-05,
+      "loss": 4.0121,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577419001901907e-05,
+      "loss": 3.993,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576580407150855e-05,
+      "loss": 4.0169,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575741812399803e-05,
+      "loss": 4.0075,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574903217648751e-05,
+      "loss": 3.9814,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574064622897699e-05,
+      "loss": 4.0154,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573226028146647e-05,
+      "loss": 3.9949,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572387433395595e-05,
+      "loss": 3.9882,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571548838644543e-05,
+      "loss": 4.0035,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570710243893491e-05,
+      "loss": 4.0069,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5698732870228116e-05,
+      "loss": 3.995,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5690346922717596e-05,
+      "loss": 4.0146,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5681960975207076e-05,
+      "loss": 3.9827,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567357502769656e-05,
+      "loss": 3.9939,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566520545898977e-05,
+      "loss": 4.0103,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565681951147925e-05,
+      "loss": 4.0101,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564843356396873e-05,
+      "loss": 4.0103,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564004761645821e-05,
+      "loss": 4.007,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563167804775142e-05,
+      "loss": 4.0129,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56232921002409e-05,
+      "loss": 3.9998,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561492253153411e-05,
+      "loss": 4.0127,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560653658402359e-05,
+      "loss": 3.9967,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559815063651307e-05,
+      "loss": 3.998,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558976468900255e-05,
+      "loss": 3.9993,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558137874149203e-05,
+      "loss": 4.0046,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5572992793981517e-05,
+      "loss": 4.0117,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5564606846470997e-05,
+      "loss": 4.0148,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5556237277764206e-05,
+      "loss": 4.0002,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5547851330253686e-05,
+      "loss": 3.9924,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5539481761546895e-05,
+      "loss": 4.0046,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5531095814036375e-05,
+      "loss": 4.0125,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5522709866525855e-05,
+      "loss": 4.0004,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5514323919015335e-05,
+      "loss": 4.0118,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5505954350308544e-05,
+      "loss": 4.0101,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5497568402798024e-05,
+      "loss": 4.0081,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5489182455287504e-05,
+      "loss": 4.002,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548081288658072e-05,
+      "loss": 3.997,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54724269390702e-05,
+      "loss": 4.0125,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546404099155968e-05,
+      "loss": 4.0103,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545565504404916e-05,
+      "loss": 4.0036,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544726909653864e-05,
+      "loss": 4.0023,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543888314902812e-05,
+      "loss": 3.9883,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54304972015176e-05,
+      "loss": 4.0024,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542211125400708e-05,
+      "loss": 4.0139,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541372530649655e-05,
+      "loss": 4.0021,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540533935898603e-05,
+      "loss": 3.999,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539695341147551e-05,
+      "loss": 3.9986,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538856746396499e-05,
+      "loss": 3.9985,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538019789525821e-05,
+      "loss": 3.9998,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537181194774769e-05,
+      "loss": 3.9975,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5363442379040904e-05,
+      "loss": 3.9938,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535505643153038e-05,
+      "loss": 4.0005,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534667048401986e-05,
+      "loss": 4.0038,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533828453650934e-05,
+      "loss": 3.9992,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532989858899882e-05,
+      "loss": 3.9989,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53215126414883e-05,
+      "loss": 3.9936,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531312669397778e-05,
+      "loss": 4.0054,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5304757125270986e-05,
+      "loss": 3.999,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296371177760466e-05,
+      "loss": 4.0105,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5287985230249946e-05,
+      "loss": 4.006,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5279599282739426e-05,
+      "loss": 3.9996,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527124609283637e-05,
+      "loss": 4.0039,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526286014532585e-05,
+      "loss": 3.9971,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525447419781533e-05,
+      "loss": 3.9999,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524608825030481e-05,
+      "loss": 4.0055,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523770230279429e-05,
+      "loss": 3.9931,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522931635528377e-05,
+      "loss": 3.9917,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522093040777325e-05,
+      "loss": 4.002,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521254446026273e-05,
+      "loss": 3.9999,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520417489155594e-05,
+      "loss": 3.9952,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519578894404542e-05,
+      "loss": 3.9883,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518741937533863e-05,
+      "loss": 3.9969,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517903342782811e-05,
+      "loss": 3.9927,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5170647480317595e-05,
+      "loss": 4.0124,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5162261532807075e-05,
+      "loss": 3.998,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5153875585296555e-05,
+      "loss": 4.0054,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145489637786035e-05,
+      "loss": 3.9913,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137103690275515e-05,
+      "loss": 4.0047,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5128717742764995e-05,
+      "loss": 4.0044,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5120331795254475e-05,
+      "loss": 4.0057,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5111945847743955e-05,
+      "loss": 4.0095,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5103559900233435e-05,
+      "loss": 4.0028,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509517395272291e-05,
+      "loss": 4.002,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5086804384016124e-05,
+      "loss": 4.0072,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507843481530933e-05,
+      "loss": 3.9921,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507004886779881e-05,
+      "loss": 3.9999,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506166292028829e-05,
+      "loss": 3.9967,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505327697277778e-05,
+      "loss": 3.9972,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504489102526726e-05,
+      "loss": 3.99,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503652145656047e-05,
+      "loss": 3.9958,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502813550904995e-05,
+      "loss": 3.9928,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501974956153943e-05,
+      "loss": 3.9976,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501136361402891e-05,
+      "loss": 4.0052,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.041686058044434,
+      "eval_runtime": 295.6276,
+      "eval_samples_per_second": 1290.783,
+      "eval_steps_per_second": 40.338,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500297766651838e-05,
+      "loss": 4.006,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499459171900786e-05,
+      "loss": 3.9874,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498620577149734e-05,
+      "loss": 3.9897,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497781982398682e-05,
+      "loss": 4.0031,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49694338764763e-05,
+      "loss": 3.9972,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496104792896578e-05,
+      "loss": 3.995,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495266198145526e-05,
+      "loss": 4.0009,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494427603394475e-05,
+      "loss": 3.9905,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493589008643423e-05,
+      "loss": 3.9925,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492750413892371e-05,
+      "loss": 3.9999,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491911819141319e-05,
+      "loss": 4.003,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491073224390267e-05,
+      "loss": 3.9927,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490234629639215e-05,
+      "loss": 4.0051,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489396034888163e-05,
+      "loss": 3.9759,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48855744013711e-05,
+      "loss": 3.9965,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487718845386058e-05,
+      "loss": 3.9802,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486880250635006e-05,
+      "loss": 3.9963,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486041655883954e-05,
+      "loss": 3.9997,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485203061132902e-05,
+      "loss": 3.983,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48436446638185e-05,
+      "loss": 4.0002,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483527509511172e-05,
+      "loss": 4.0123,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48268891476012e-05,
+      "loss": 3.9949,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481850320009068e-05,
+      "loss": 4.002,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481011725258016e-05,
+      "loss": 4.0028,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480173130506964e-05,
+      "loss": 3.9994,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4793361736362846e-05,
+      "loss": 3.9957,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4784992167656055e-05,
+      "loss": 3.9862,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4776606220145535e-05,
+      "loss": 3.997,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4768220272635015e-05,
+      "loss": 3.9903,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4759834325124495e-05,
+      "loss": 3.9919,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4751448377613975e-05,
+      "loss": 4.0015,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4743078808907184e-05,
+      "loss": 3.9998,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4734692861396664e-05,
+      "loss": 3.9963,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472630691388615e-05,
+      "loss": 4.0063,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471792096637563e-05,
+      "loss": 3.9895,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470953501886511e-05,
+      "loss": 3.9969,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470114907135459e-05,
+      "loss": 3.9954,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469276312384407e-05,
+      "loss": 3.9797,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468437717633355e-05,
+      "loss": 3.9892,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467600760762676e-05,
+      "loss": 3.9885,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466762166011624e-05,
+      "loss": 3.9842,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465923571260572e-05,
+      "loss": 3.9943,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46508497650952e-05,
+      "loss": 3.9956,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464248019638841e-05,
+      "loss": 3.9924,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463409424887789e-05,
+      "loss": 3.9957,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462570830136737e-05,
+      "loss": 3.9994,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4617338732660584e-05,
+      "loss": 3.9861,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4608952785150064e-05,
+      "loss": 4.0001,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4600566837639544e-05,
+      "loss": 3.977,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4592180890129024e-05,
+      "loss": 3.9947,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4583794942618504e-05,
+      "loss": 3.9866,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575408995107984e-05,
+      "loss": 4.0027,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456702304759746e-05,
+      "loss": 3.9862,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455863710008694e-05,
+      "loss": 3.9827,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455026753138015e-05,
+      "loss": 3.9887,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454188158386963e-05,
+      "loss": 3.9893,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533495636359106e-05,
+      "loss": 3.9985,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452512606765232e-05,
+      "loss": 3.9715,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45167401201418e-05,
+      "loss": 4.0028,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450835417263129e-05,
+      "loss": 3.991,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449996822512076e-05,
+      "loss": 3.9684,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449158227761024e-05,
+      "loss": 3.9994,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448319633009972e-05,
+      "loss": 3.9784,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44748103825892e-05,
+      "loss": 3.973,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446642443507868e-05,
+      "loss": 3.9842,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445805486637189e-05,
+      "loss": 3.9927,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4449685297665107e-05,
+      "loss": 3.9848,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444129935015458e-05,
+      "loss": 3.9913,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443291340264406e-05,
+      "loss": 3.9727,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442452745513354e-05,
+      "loss": 3.9764,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4416141507623026e-05,
+      "loss": 3.9938,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4407755560112506e-05,
+      "loss": 3.9959,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4399369612601986e-05,
+      "loss": 3.9935,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4391000043895195e-05,
+      "loss": 3.9909,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4382614096384675e-05,
+      "loss": 3.9972,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4374228148874155e-05,
+      "loss": 3.9838,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4365842201363635e-05,
+      "loss": 3.9909,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4357456253853115e-05,
+      "loss": 3.9836,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4349070306342595e-05,
+      "loss": 3.9832,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340684358832075e-05,
+      "loss": 3.9869,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4332298411321555e-05,
+      "loss": 3.9803,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4323928842614764e-05,
+      "loss": 3.9975,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315542895104244e-05,
+      "loss": 3.9972,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307156947593724e-05,
+      "loss": 3.9878,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429877100008321e-05,
+      "loss": 3.9708,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429040143137642e-05,
+      "loss": 3.9892,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42820154838659e-05,
+      "loss": 3.9935,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427362953635538e-05,
+      "loss": 3.9872,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426524358884486e-05,
+      "loss": 3.994,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425687402013807e-05,
+      "loss": 3.9976,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424848807262755e-05,
+      "loss": 3.9882,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424010212511703e-05,
+      "loss": 3.9877,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423171617760651e-05,
+      "loss": 3.9825,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422333023009599e-05,
+      "loss": 3.9958,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421494428258547e-05,
+      "loss": 3.9914,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420655833507495e-05,
+      "loss": 3.9911,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419817238756443e-05,
+      "loss": 3.9852,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4189802818857644e-05,
+      "loss": 3.9738,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4181416871347124e-05,
+      "loss": 3.9862,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41730309238366e-05,
+      "loss": 3.998,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416464497632608e-05,
+      "loss": 3.991,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415625902881556e-05,
+      "loss": 3.977,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147889460108767e-05,
+      "loss": 3.988,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139503512598247e-05,
+      "loss": 3.9806,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131117565087726e-05,
+      "loss": 3.9845,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122731617577206e-05,
+      "loss": 3.9827,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114362048870416e-05,
+      "loss": 3.9784,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4105976101359895e-05,
+      "loss": 3.9831,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409759015384938e-05,
+      "loss": 3.9869,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408920420633886e-05,
+      "loss": 3.9847,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408083463763207e-05,
+      "loss": 3.9781,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407244869012155e-05,
+      "loss": 3.9785,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406406274261103e-05,
+      "loss": 3.99,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405567679510051e-05,
+      "loss": 3.9855,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404729084758999e-05,
+      "loss": 3.9924,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40389212788832e-05,
+      "loss": 3.9881,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403053533137268e-05,
+      "loss": 3.9897,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402216576266589e-05,
+      "loss": 3.9889,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401377981515537e-05,
+      "loss": 3.9792,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4005410246448585e-05,
+      "loss": 3.9872,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3997024298938065e-05,
+      "loss": 3.9898,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3988638351427545e-05,
+      "loss": 3.9771,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3980252403917025e-05,
+      "loss": 3.9832,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3971866456406505e-05,
+      "loss": 3.9797,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3963480508895985e-05,
+      "loss": 3.9857,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3955094561385465e-05,
+      "loss": 3.9835,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3946708613874945e-05,
+      "loss": 3.9733,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3938322666364425e-05,
+      "loss": 3.9763,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929936718853905e-05,
+      "loss": 3.9786,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921550771343385e-05,
+      "loss": 3.9961,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913164823832865e-05,
+      "loss": 3.9847,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904795255126074e-05,
+      "loss": 3.986,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896409307615554e-05,
+      "loss": 3.9767,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3888023360105034e-05,
+      "loss": 3.9884,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387963741259452e-05,
+      "loss": 3.9894,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871251465084e-05,
+      "loss": 3.9874,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386286551757348e-05,
+      "loss": 3.9926,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854479570062953e-05,
+      "loss": 3.9842,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384611000135617e-05,
+      "loss": 3.9904,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383772405384565e-05,
+      "loss": 3.9934,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382933810633513e-05,
+      "loss": 3.975,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38209521588246e-05,
+      "loss": 3.986,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381256621131408e-05,
+      "loss": 3.9806,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380418026380356e-05,
+      "loss": 3.9835,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379581069509678e-05,
+      "loss": 3.9762,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378742474758626e-05,
+      "loss": 3.9795,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377903880007574e-05,
+      "loss": 3.9764,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377065285256522e-05,
+      "loss": 3.9828,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37622669050547e-05,
+      "loss": 3.988,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.032730579376221,
+      "eval_runtime": 292.5381,
+      "eval_samples_per_second": 1304.415,
+      "eval_steps_per_second": 40.764,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375388095754418e-05,
+      "loss": 3.985,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374549501003366e-05,
+      "loss": 3.9715,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373710906252314e-05,
+      "loss": 3.975,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372877225142381e-05,
+      "loss": 3.9867,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372038630391329e-05,
+      "loss": 3.98,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371200035640277e-05,
+      "loss": 3.9819,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370361440889225e-05,
+      "loss": 3.9822,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3695228461381725e-05,
+      "loss": 3.9748,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368684251387121e-05,
+      "loss": 3.976,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367847294516443e-05,
+      "loss": 3.9874,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36700869976539e-05,
+      "loss": 3.9835,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366170105014338e-05,
+      "loss": 3.9803,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365331510263286e-05,
+      "loss": 3.993,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364492915512234e-05,
+      "loss": 3.9588,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363654320761182e-05,
+      "loss": 3.9798,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36281572601013e-05,
+      "loss": 3.9645,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361977131259078e-05,
+      "loss": 3.9847,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361138536508026e-05,
+      "loss": 3.9822,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360299941756974e-05,
+      "loss": 3.9721,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359461347005922e-05,
+      "loss": 3.9772,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358624390135243e-05,
+      "loss": 4.0013,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357785795384191e-05,
+      "loss": 3.9816,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356947200633139e-05,
+      "loss": 3.9862,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3561086058820876e-05,
+      "loss": 3.9886,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3552700111310356e-05,
+      "loss": 3.9876,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3544330542603565e-05,
+      "loss": 3.9771,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3535944595093045e-05,
+      "loss": 3.9738,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3527558647582525e-05,
+      "loss": 3.9806,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3519172700072005e-05,
+      "loss": 3.977,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3510803131365214e-05,
+      "loss": 3.978,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3502417183854694e-05,
+      "loss": 3.9863,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3494031236344174e-05,
+      "loss": 3.9874,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3485645288833654e-05,
+      "loss": 3.9829,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347727572012686e-05,
+      "loss": 3.9887,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346888977261634e-05,
+      "loss": 3.9752,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346052020390956e-05,
+      "loss": 3.9837,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345213425639904e-05,
+      "loss": 3.9822,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.344374830888852e-05,
+      "loss": 3.9624,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3435362361378e-05,
+      "loss": 3.9795,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342697641386748e-05,
+      "loss": 3.9703,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341859046635696e-05,
+      "loss": 3.9697,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341020451884644e-05,
+      "loss": 3.9803,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340181857133591e-05,
+      "loss": 3.9869,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339344900262913e-05,
+      "loss": 3.9705,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338506305511861e-05,
+      "loss": 3.9799,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337667710760808e-05,
+      "loss": 3.9821,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336829116009757e-05,
+      "loss": 3.9721,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3359921591390784e-05,
+      "loss": 3.9826,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335155202268399e-05,
+      "loss": 3.9635,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334316607517347e-05,
+      "loss": 3.9772,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333478012766295e-05,
+      "loss": 3.975,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332639418015243e-05,
+      "loss": 3.9871,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331800823264191e-05,
+      "loss": 3.9742,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309622285131386e-05,
+      "loss": 3.967,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3301236337620866e-05,
+      "loss": 3.9716,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.329286676891408e-05,
+      "loss": 3.9731,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284480821403555e-05,
+      "loss": 3.9865,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3276094873893035e-05,
+      "loss": 3.9579,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326770892638252e-05,
+      "loss": 3.9851,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3259322978872e-05,
+      "loss": 3.9747,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325093703136148e-05,
+      "loss": 3.9584,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324255108385096e-05,
+      "loss": 3.9791,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323416513634044e-05,
+      "loss": 3.966,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322579556763365e-05,
+      "loss": 3.96,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321740962012313e-05,
+      "loss": 3.9663,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320902367261261e-05,
+      "loss": 3.9775,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320063772510209e-05,
+      "loss": 3.9734,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319228453519903e-05,
+      "loss": 3.9689,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318389858768851e-05,
+      "loss": 3.9647,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317551264017799e-05,
+      "loss": 3.9617,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3167126692667475e-05,
+      "loss": 3.9748,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3158740745156955e-05,
+      "loss": 3.9808,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3150371176450164e-05,
+      "loss": 3.9775,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3141985228939644e-05,
+      "loss": 3.9831,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3133599281429124e-05,
+      "loss": 3.9784,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3125213333918604e-05,
+      "loss": 3.9688,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3116827386408084e-05,
+      "loss": 3.9748,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3108441438897564e-05,
+      "loss": 3.9708,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3100055491387044e-05,
+      "loss": 3.9732,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3091669543876524e-05,
+      "loss": 3.9654,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308329997516973e-05,
+      "loss": 3.9697,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.307491402765921e-05,
+      "loss": 3.9854,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306654445895243e-05,
+      "loss": 3.9802,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305815851144191e-05,
+      "loss": 3.9769,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304977256393139e-05,
+      "loss": 3.9564,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304138661642087e-05,
+      "loss": 3.9736,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303300066891035e-05,
+      "loss": 3.9781,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302461472139983e-05,
+      "loss": 3.9758,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301622877388931e-05,
+      "loss": 3.9765,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.300784282637879e-05,
+      "loss": 3.9862,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299945687886827e-05,
+      "loss": 3.9733,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299107093135775e-05,
+      "loss": 3.9717,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298268498384722e-05,
+      "loss": 3.9685,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297431541514044e-05,
+      "loss": 3.9785,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296592946762992e-05,
+      "loss": 3.9768,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29575435201194e-05,
+      "loss": 3.9807,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294915757260888e-05,
+      "loss": 3.9672,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294077162509836e-05,
+      "loss": 3.9629,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293240205639157e-05,
+      "loss": 3.9692,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2924016108881046e-05,
+      "loss": 3.9822,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2915630161370526e-05,
+      "loss": 3.9764,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2907244213860006e-05,
+      "loss": 3.9647,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289887464515322e-05,
+      "loss": 3.9724,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2890488697642695e-05,
+      "loss": 3.964,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2882102750132175e-05,
+      "loss": 3.971,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2873716802621655e-05,
+      "loss": 3.9662,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2865347233914864e-05,
+      "loss": 3.9671,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2856961286404344e-05,
+      "loss": 3.9718,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284857533889383e-05,
+      "loss": 3.9732,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284018939138331e-05,
+      "loss": 3.967,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283181982267652e-05,
+      "loss": 3.9661,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2823433875166e-05,
+      "loss": 3.9668,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281504792765548e-05,
+      "loss": 3.9761,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280666198014496e-05,
+      "loss": 3.9725,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279827603263444e-05,
+      "loss": 3.98,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278990646392765e-05,
+      "loss": 3.9714,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2781536895220865e-05,
+      "loss": 3.9765,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277315094771034e-05,
+      "loss": 3.9714,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276476500019982e-05,
+      "loss": 3.9636,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27563790526893e-05,
+      "loss": 3.9764,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2747993105178784e-05,
+      "loss": 3.9731,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2739607157668264e-05,
+      "loss": 3.9637,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2731237588961474e-05,
+      "loss": 3.9679,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2722851641450954e-05,
+      "loss": 3.9692,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2714465693940433e-05,
+      "loss": 3.9711,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2706079746429913e-05,
+      "loss": 3.9629,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269769379891939e-05,
+      "loss": 3.962,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268930785140887e-05,
+      "loss": 3.964,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268093828270208e-05,
+      "loss": 3.9662,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267255233519156e-05,
+      "loss": 3.9812,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266416638768104e-05,
+      "loss": 3.9743,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265578044017052e-05,
+      "loss": 3.9744,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264739449266e-05,
+      "loss": 3.963,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263902492395322e-05,
+      "loss": 3.9765,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26306389764427e-05,
+      "loss": 3.9753,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262225302893218e-05,
+      "loss": 3.9767,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261386708142166e-05,
+      "loss": 3.9751,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260548113391114e-05,
+      "loss": 3.9714,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259711156520435e-05,
+      "loss": 3.978,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258872561769383e-05,
+      "loss": 3.9789,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258033967018331e-05,
+      "loss": 3.9628,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257195372267279e-05,
+      "loss": 3.9687,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.256356777516227e-05,
+      "loss": 3.9706,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255518182765175e-05,
+      "loss": 3.9685,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254679588014123e-05,
+      "loss": 3.962,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253840993263071e-05,
+      "loss": 3.9681,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253004036392392e-05,
+      "loss": 3.9593,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25216544164134e-05,
+      "loss": 3.9767,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251326846890288e-05,
+      "loss": 3.9675,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.024904727935791,
+      "eval_runtime": 295.9734,
+      "eval_samples_per_second": 1289.274,
+      "eval_steps_per_second": 40.291,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2504882521392356e-05,
+      "loss": 3.98,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2496496573881836e-05,
+      "loss": 3.9586,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2488110626371316e-05,
+      "loss": 3.9633,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2479724678860796e-05,
+      "loss": 3.9753,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2471338731350276e-05,
+      "loss": 3.9688,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2462952783839755e-05,
+      "loss": 3.9672,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2454566836329235e-05,
+      "loss": 3.9652,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2446180888818715e-05,
+      "loss": 3.962,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2437794941308195e-05,
+      "loss": 3.9648,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2429408993797675e-05,
+      "loss": 3.976,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2421023046287155e-05,
+      "loss": 3.9711,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2412637098776635e-05,
+      "loss": 3.9681,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2404251151266115e-05,
+      "loss": 3.9755,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2395865203755595e-05,
+      "loss": 3.9496,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2387479256245075e-05,
+      "loss": 3.9634,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2379093308734555e-05,
+      "loss": 3.9564,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2370707361224035e-05,
+      "loss": 3.9686,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2362321413713515e-05,
+      "loss": 3.965,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2353935466202995e-05,
+      "loss": 3.9626,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2345549518692475e-05,
+      "loss": 3.9656,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2337179949985684e-05,
+      "loss": 3.9856,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2328794002475164e-05,
+      "loss": 3.97,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2320408054964644e-05,
+      "loss": 3.9699,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2312022107454124e-05,
+      "loss": 3.9732,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230365253874734e-05,
+      "loss": 3.9791,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229526659123682e-05,
+      "loss": 3.9586,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22868806437263e-05,
+      "loss": 3.9617,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227849469621578e-05,
+      "loss": 3.9722,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227012512750899e-05,
+      "loss": 3.9579,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226173917999847e-05,
+      "loss": 3.9654,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225335323248795e-05,
+      "loss": 3.9723,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224498366378116e-05,
+      "loss": 3.973,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223659771627064e-05,
+      "loss": 3.9716,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222822814756385e-05,
+      "loss": 3.9749,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221984220005333e-05,
+      "loss": 3.9627,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221145625254281e-05,
+      "loss": 3.9726,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220307030503229e-05,
+      "loss": 3.9666,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219468435752177e-05,
+      "loss": 3.9505,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218629841001125e-05,
+      "loss": 3.9688,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217791246250073e-05,
+      "loss": 3.9553,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216952651499021e-05,
+      "loss": 3.9561,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216114056747969e-05,
+      "loss": 3.9661,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21527709987729e-05,
+      "loss": 3.9722,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214438505126238e-05,
+      "loss": 3.9541,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213599910375186e-05,
+      "loss": 3.9735,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212761315624134e-05,
+      "loss": 3.9661,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211922720873082e-05,
+      "loss": 3.9649,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21108412612203e-05,
+      "loss": 3.9693,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2102455313709775e-05,
+      "loss": 3.9491,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.209406936619926e-05,
+      "loss": 3.9635,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208569979749248e-05,
+      "loss": 3.9625,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207731384998196e-05,
+      "loss": 3.9691,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206892790247143e-05,
+      "loss": 3.9648,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206054195496091e-05,
+      "loss": 3.9504,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205217238625413e-05,
+      "loss": 3.9585,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20437864387436e-05,
+      "loss": 3.9592,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2035416870036816e-05,
+      "loss": 3.9787,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2027030922526296e-05,
+      "loss": 3.9445,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2018644975015776e-05,
+      "loss": 3.9709,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201025902750525e-05,
+      "loss": 3.964,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200187307999473e-05,
+      "loss": 3.9468,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1993487132484216e-05,
+      "loss": 3.9631,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198511756377743e-05,
+      "loss": 3.9586,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1976731616266905e-05,
+      "loss": 3.9472,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1968345668756385e-05,
+      "loss": 3.9489,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1959959721245865e-05,
+      "loss": 3.9708,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1951573773735344e-05,
+      "loss": 3.9596,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1943187826224824e-05,
+      "loss": 3.9553,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1934818257518034e-05,
+      "loss": 3.9484,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1926432310007514e-05,
+      "loss": 3.9512,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1918046362496993e-05,
+      "loss": 3.9627,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1909660414986473e-05,
+      "loss": 3.9679,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190127446747595e-05,
+      "loss": 3.963,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189288851996543e-05,
+      "loss": 3.9695,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188450257245491e-05,
+      "loss": 3.966,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187611662494439e-05,
+      "loss": 3.9641,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186774705623761e-05,
+      "loss": 3.959,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185936110872709e-05,
+      "loss": 3.9543,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185097516121657e-05,
+      "loss": 3.9576,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184258921370605e-05,
+      "loss": 3.9582,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183420326619553e-05,
+      "loss": 3.9569,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182581731868501e-05,
+      "loss": 3.9675,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181744774997822e-05,
+      "loss": 3.9734,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18090618024677e-05,
+      "loss": 3.9627,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180067585495718e-05,
+      "loss": 3.9443,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179228990744666e-05,
+      "loss": 3.9617,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178390395993614e-05,
+      "loss": 3.9659,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177551801242562e-05,
+      "loss": 3.966,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.17671320649151e-05,
+      "loss": 3.959,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1758762496208314e-05,
+      "loss": 3.9739,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1750376548697794e-05,
+      "loss": 3.962,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174199060118727e-05,
+      "loss": 3.9584,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173360465367675e-05,
+      "loss": 3.9581,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172521870616623e-05,
+      "loss": 3.9682,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1716832758655707e-05,
+      "loss": 3.9612,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1708446811145187e-05,
+      "loss": 3.9678,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1700060863634667e-05,
+      "loss": 3.9555,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1691691294927876e-05,
+      "loss": 3.9517,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1683305347417356e-05,
+      "loss": 3.9581,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1674919399906836e-05,
+      "loss": 3.9705,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1666533452396315e-05,
+      "loss": 3.9646,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165816388368953e-05,
+      "loss": 3.953,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164977793617901e-05,
+      "loss": 3.9581,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164139198866849e-05,
+      "loss": 3.9516,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163300604115797e-05,
+      "loss": 3.9631,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162462009364745e-05,
+      "loss": 3.9534,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161625052494066e-05,
+      "loss": 3.9523,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160786457743014e-05,
+      "loss": 3.9574,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159947862991962e-05,
+      "loss": 3.96,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15910926824091e-05,
+      "loss": 3.959,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158272311370231e-05,
+      "loss": 3.9548,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1574353544995525e-05,
+      "loss": 3.951,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1565967597485005e-05,
+      "loss": 3.9614,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1557581649974485e-05,
+      "loss": 3.9621,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1549195702463965e-05,
+      "loss": 3.9592,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1540809754953445e-05,
+      "loss": 3.9636,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1532423807442925e-05,
+      "loss": 3.9648,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1524054238736134e-05,
+      "loss": 3.958,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1515668291225614e-05,
+      "loss": 3.9517,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1507282343715094e-05,
+      "loss": 3.9628,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1498896396204574e-05,
+      "loss": 3.958,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490510448694054e-05,
+      "loss": 3.9517,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1482124501183534e-05,
+      "loss": 3.9576,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1473738553673014e-05,
+      "loss": 3.9556,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146536898496622e-05,
+      "loss": 3.962,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145698303745571e-05,
+      "loss": 3.9471,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144859708994519e-05,
+      "loss": 3.9559,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144021114243467e-05,
+      "loss": 3.9528,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143182519492415e-05,
+      "loss": 3.9528,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142343924741362e-05,
+      "loss": 3.9627,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14150532999031e-05,
+      "loss": 3.9652,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140668373119632e-05,
+      "loss": 3.9568,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13982977836858e-05,
+      "loss": 3.9507,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138991183617527e-05,
+      "loss": 3.966,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138152588866475e-05,
+      "loss": 3.9617,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137313994115423e-05,
+      "loss": 3.9616,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136477037244745e-05,
+      "loss": 3.9658,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135638442493693e-05,
+      "loss": 3.9599,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134799847742641e-05,
+      "loss": 3.9694,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133961252991589e-05,
+      "loss": 3.9654,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1331242961209096e-05,
+      "loss": 3.9511,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1322857013698576e-05,
+      "loss": 3.9564,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1314471066188056e-05,
+      "loss": 3.956,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1306085118677536e-05,
+      "loss": 3.957,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1297699171167016e-05,
+      "loss": 3.9534,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1289313223656496e-05,
+      "loss": 3.9563,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1280927276145976e-05,
+      "loss": 3.9496,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1272541328635456e-05,
+      "loss": 3.9648,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1264171759928665e-05,
+      "loss": 3.9565,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.018359184265137,
+      "eval_runtime": 300.9569,
+      "eval_samples_per_second": 1267.926,
+      "eval_steps_per_second": 39.624,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1255785812418145e-05,
+      "loss": 3.9572,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1247399864907625e-05,
+      "loss": 3.9533,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123901391739711e-05,
+      "loss": 3.947,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123064434869032e-05,
+      "loss": 3.9624,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.12222584011798e-05,
+      "loss": 3.9577,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121387245366928e-05,
+      "loss": 3.9573,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120548650615876e-05,
+      "loss": 3.9514,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119711693745197e-05,
+      "loss": 3.9472,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118873098994145e-05,
+      "loss": 3.9578,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118034504243093e-05,
+      "loss": 3.958,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117195909492041e-05,
+      "loss": 3.9604,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116358952621362e-05,
+      "loss": 3.9542,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11552035787031e-05,
+      "loss": 3.9687,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114681763119258e-05,
+      "loss": 3.9369,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1138431683682065e-05,
+      "loss": 3.952,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1130062114975274e-05,
+      "loss": 3.9446,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1121676167464754e-05,
+      "loss": 3.9553,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1113290219954234e-05,
+      "loss": 3.9561,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1104904272443714e-05,
+      "loss": 3.951,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1096534703736923e-05,
+      "loss": 3.9527,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108816513503013e-05,
+      "loss": 3.9752,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107977918751961e-05,
+      "loss": 3.9581,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107139324000909e-05,
+      "loss": 3.9554,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.106300729249857e-05,
+      "loss": 3.9645,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105462134498805e-05,
+      "loss": 3.9654,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.104623539747753e-05,
+      "loss": 3.9515,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103784944996702e-05,
+      "loss": 3.9476,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102947988126023e-05,
+      "loss": 3.9606,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102109393374971e-05,
+      "loss": 3.946,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.101270798623919e-05,
+      "loss": 3.9554,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.100432203872867e-05,
+      "loss": 3.9549,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099595247002188e-05,
+      "loss": 3.9649,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098756652251136e-05,
+      "loss": 3.9607,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097918057500084e-05,
+      "loss": 3.9577,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097079462749032e-05,
+      "loss": 3.9546,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0962425058783526e-05,
+      "loss": 3.9626,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0954039111273006e-05,
+      "loss": 3.9562,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0945653163762486e-05,
+      "loss": 3.9389,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09372835950557e-05,
+      "loss": 3.9577,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092889764754518e-05,
+      "loss": 3.9443,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092051170003466e-05,
+      "loss": 3.9493,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091212575252414e-05,
+      "loss": 3.9515,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090373980501362e-05,
+      "loss": 3.96,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08953538575031e-05,
+      "loss": 3.9485,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088696790999258e-05,
+      "loss": 3.953,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0878581962482055e-05,
+      "loss": 3.9609,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0870196014971535e-05,
+      "loss": 3.9476,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0861810067461015e-05,
+      "loss": 3.9631,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0853424119950495e-05,
+      "loss": 3.9399,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0845038172439975e-05,
+      "loss": 3.9479,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083666860373319e-05,
+      "loss": 3.9549,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082828265622267e-05,
+      "loss": 3.9564,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081989670871215e-05,
+      "loss": 3.9554,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081151076120163e-05,
+      "loss": 3.9389,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.080312481369111e-05,
+      "loss": 3.9526,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.079475524498432e-05,
+      "loss": 3.9441,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07863692974738e-05,
+      "loss": 3.967,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077798334996328e-05,
+      "loss": 3.9306,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076959740245276e-05,
+      "loss": 3.9602,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076121145494224e-05,
+      "loss": 3.9523,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075282550743172e-05,
+      "loss": 3.9355,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07444395599212e-05,
+      "loss": 3.9529,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073606999121441e-05,
+      "loss": 3.9486,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072768404370389e-05,
+      "loss": 3.9322,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0719298096193375e-05,
+      "loss": 3.9379,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0710912148682855e-05,
+      "loss": 3.9605,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0702526201172335e-05,
+      "loss": 3.9481,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0694140253661815e-05,
+      "loss": 3.9457,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0685754306151295e-05,
+      "loss": 3.9379,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067736835864077e-05,
+      "loss": 3.9384,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066898241113025e-05,
+      "loss": 3.9526,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0660612842423464e-05,
+      "loss": 3.9558,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0652226894912944e-05,
+      "loss": 3.9498,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.064384094740242e-05,
+      "loss": 3.9592,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06354549998919e-05,
+      "loss": 3.9562,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062708543118511e-05,
+      "loss": 3.9492,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061871586247833e-05,
+      "loss": 3.9487,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061032991496781e-05,
+      "loss": 3.9455,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060194396745729e-05,
+      "loss": 3.949,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059355801994677e-05,
+      "loss": 3.9492,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058517207243624e-05,
+      "loss": 3.944,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057680250372946e-05,
+      "loss": 3.9571,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056841655621894e-05,
+      "loss": 3.956,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056003060870842e-05,
+      "loss": 3.9521,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055164466119789e-05,
+      "loss": 3.9356,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0543275092491106e-05,
+      "loss": 3.9453,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0534889144980586e-05,
+      "loss": 3.9598,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0526503197470066e-05,
+      "loss": 3.9545,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0518117249959546e-05,
+      "loss": 3.9484,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509731302449023e-05,
+      "loss": 3.9605,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.050136173374224e-05,
+      "loss": 3.9513,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0492975786231715e-05,
+      "loss": 3.9467,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0484589838721195e-05,
+      "loss": 3.9477,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.047622027001441e-05,
+      "loss": 3.956,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046783432250389e-05,
+      "loss": 3.9482,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0459448374993364e-05,
+      "loss": 3.9615,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0451062427482848e-05,
+      "loss": 3.9396,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0442676479972328e-05,
+      "loss": 3.9477,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0434290532461808e-05,
+      "loss": 3.9452,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0425904584951288e-05,
+      "loss": 3.956,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0417518637440767e-05,
+      "loss": 3.9596,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0409132689930247e-05,
+      "loss": 3.9388,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0400746742419727e-05,
+      "loss": 3.9488,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0392360794909207e-05,
+      "loss": 3.9398,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0383974847398687e-05,
+      "loss": 3.9508,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.03756052786919e-05,
+      "loss": 3.935,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.036721933118138e-05,
+      "loss": 3.9473,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035883338367086e-05,
+      "loss": 3.9492,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035044743616034e-05,
+      "loss": 3.9447,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.034207786745355e-05,
+      "loss": 3.9505,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.033369191994303e-05,
+      "loss": 3.9476,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0325305972432512e-05,
+      "loss": 3.9355,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.031693640372572e-05,
+      "loss": 3.9539,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.03085504562152e-05,
+      "loss": 3.9518,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.030016450870468e-05,
+      "loss": 3.9487,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.029177856119416e-05,
+      "loss": 3.9533,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.028339261368364e-05,
+      "loss": 3.9548,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0275023044976853e-05,
+      "loss": 3.9493,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0266637097466333e-05,
+      "loss": 3.9395,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0258251149955813e-05,
+      "loss": 3.9503,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0249865202445293e-05,
+      "loss": 3.9454,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0241479254934773e-05,
+      "loss": 3.9419,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.023309330742425e-05,
+      "loss": 3.9442,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.022470735991373e-05,
+      "loss": 3.9464,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.021632141240321e-05,
+      "loss": 3.9495,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0207951843696426e-05,
+      "loss": 3.9393,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01995658961859e-05,
+      "loss": 3.9439,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0191196327479115e-05,
+      "loss": 3.9415,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0182810379968595e-05,
+      "loss": 3.9474,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0174424432458075e-05,
+      "loss": 3.9472,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.016603848494755e-05,
+      "loss": 3.9529,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0157668916240767e-05,
+      "loss": 3.9475,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0149282968730247e-05,
+      "loss": 3.9423,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.014089702121972e-05,
+      "loss": 3.9513,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0132511073709203e-05,
+      "loss": 3.9521,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0124125126198683e-05,
+      "loss": 3.9543,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115739178688163e-05,
+      "loss": 3.9551,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0107353231177643e-05,
+      "loss": 3.9471,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098983662470852e-05,
+      "loss": 3.9601,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0090597714960332e-05,
+      "loss": 3.9521,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0082228146253548e-05,
+      "loss": 3.9428,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073842198743025e-05,
+      "loss": 3.9458,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0065456251232505e-05,
+      "loss": 3.9427,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0057070303721985e-05,
+      "loss": 3.9443,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0048684356211465e-05,
+      "loss": 3.9394,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0040298408700945e-05,
+      "loss": 3.9497,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0031912461190425e-05,
+      "loss": 3.9378,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0023526513679905e-05,
+      "loss": 3.9512,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0015140566169388e-05,
+      "loss": 3.949,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0006754618658868e-05,
+      "loss": 3.947,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.013096809387207,
+      "eval_runtime": 303.1106,
+      "eval_samples_per_second": 1258.917,
+      "eval_steps_per_second": 39.342,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998368671148348e-05,
+      "loss": 3.9394,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989982723637828e-05,
+      "loss": 3.9374,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981596776127308e-05,
+      "loss": 3.9489,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9973210828616788e-05,
+      "loss": 3.9498,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964824881106264e-05,
+      "loss": 3.947,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956438933595744e-05,
+      "loss": 3.9414,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9948052986085224e-05,
+      "loss": 3.937,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939667038574704e-05,
+      "loss": 3.94,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931281091064184e-05,
+      "loss": 3.9539,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922895143553664e-05,
+      "loss": 3.9443,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914509196043144e-05,
+      "loss": 3.9498,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9906123248532624e-05,
+      "loss": 3.9574,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9897737301022104e-05,
+      "loss": 3.9289,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889351353511584e-05,
+      "loss": 3.937,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880965406001067e-05,
+      "loss": 3.9345,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9872579458490547e-05,
+      "loss": 3.9427,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9864193510980027e-05,
+      "loss": 3.9459,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98558075634695e-05,
+      "loss": 3.9424,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984742161595898e-05,
+      "loss": 3.9375,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9839052047252196e-05,
+      "loss": 3.9627,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830666099741676e-05,
+      "loss": 3.9451,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822280152231153e-05,
+      "loss": 3.9481,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813894204720633e-05,
+      "loss": 3.9463,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.980552463601385e-05,
+      "loss": 3.9601,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.979713868850332e-05,
+      "loss": 3.9398,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788752740992805e-05,
+      "loss": 3.9398,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9780366793482285e-05,
+      "loss": 3.9468,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.97719972247755e-05,
+      "loss": 3.9367,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763611277264974e-05,
+      "loss": 3.9423,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755225329754454e-05,
+      "loss": 3.9431,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9746839382243934e-05,
+      "loss": 3.956,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9738453434733414e-05,
+      "loss": 3.9455,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9730067487222897e-05,
+      "loss": 3.9495,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9721681539712377e-05,
+      "loss": 3.9446,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9713311971005586e-05,
+      "loss": 3.9474,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9704942402298795e-05,
+      "loss": 3.9458,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9696556454788275e-05,
+      "loss": 3.9305,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968817050727776e-05,
+      "loss": 3.9417,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967978455976724e-05,
+      "loss": 3.9319,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967139861225672e-05,
+      "loss": 3.9419,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96630126647462e-05,
+      "loss": 3.9418,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.965462671723568e-05,
+      "loss": 3.9496,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.964624076972516e-05,
+      "loss": 3.9388,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963785482221464e-05,
+      "loss": 3.9433,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962946887470412e-05,
+      "loss": 3.9491,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96210829271936e-05,
+      "loss": 3.9381,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961269697968308e-05,
+      "loss": 3.9503,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.960432741097629e-05,
+      "loss": 3.9272,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.959594146346577e-05,
+      "loss": 3.9375,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.958755551595525e-05,
+      "loss": 3.9473,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957916956844473e-05,
+      "loss": 3.9443,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957078362093421e-05,
+      "loss": 3.9427,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562414052227423e-05,
+      "loss": 3.9245,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9554028104716903e-05,
+      "loss": 3.9452,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545658536010112e-05,
+      "loss": 3.93,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537272588499592e-05,
+      "loss": 3.9572,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528886640989072e-05,
+      "loss": 3.9245,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520500693478552e-05,
+      "loss": 3.9481,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9512114745968032e-05,
+      "loss": 3.9377,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503728798457515e-05,
+      "loss": 3.9271,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.949534285094699e-05,
+      "loss": 3.9432,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486973282240204e-05,
+      "loss": 3.9368,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478587334729684e-05,
+      "loss": 3.9195,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947020138721916e-05,
+      "loss": 3.9295,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.946181543970864e-05,
+      "loss": 3.9508,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.945342949219812e-05,
+      "loss": 3.9354,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.94450435446876e-05,
+      "loss": 3.9333,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.943667397598081e-05,
+      "loss": 3.9315,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942828802847029e-05,
+      "loss": 3.9275,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941990208095977e-05,
+      "loss": 3.9401,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411516133449253e-05,
+      "loss": 3.9473,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9403130185938733e-05,
+      "loss": 3.9405,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394744238428213e-05,
+      "loss": 3.951,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386358290917693e-05,
+      "loss": 3.9427,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377972343407173e-05,
+      "loss": 3.9426,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369602774700382e-05,
+      "loss": 3.9329,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9361216827189862e-05,
+      "loss": 3.9397,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9352830879679345e-05,
+      "loss": 3.941,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344444932168825e-05,
+      "loss": 3.933,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9336058984658305e-05,
+      "loss": 3.9335,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9327673037147785e-05,
+      "loss": 3.9449,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9319303468440994e-05,
+      "loss": 3.9483,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310917520930474e-05,
+      "loss": 3.9384,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302531573419954e-05,
+      "loss": 3.9237,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9294145625909438e-05,
+      "loss": 3.9397,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9285759678398917e-05,
+      "loss": 3.9478,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9277373730888397e-05,
+      "loss": 3.9468,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.926898778337787e-05,
+      "loss": 3.9365,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.926060183586735e-05,
+      "loss": 3.953,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252232267160566e-05,
+      "loss": 3.9421,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243846319650046e-05,
+      "loss": 3.9318,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.923547675094326e-05,
+      "loss": 3.9391,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.922709080343274e-05,
+      "loss": 3.9428,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921870485592222e-05,
+      "loss": 3.9358,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.92103189084117e-05,
+      "loss": 3.953,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9201932960901175e-05,
+      "loss": 3.9298,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9193547013390655e-05,
+      "loss": 3.9319,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.918517744468387e-05,
+      "loss": 3.9352,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9176791497173344e-05,
+      "loss": 3.9475,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9168405549662824e-05,
+      "loss": 3.9453,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9160019602152304e-05,
+      "loss": 3.9309,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9151633654641784e-05,
+      "loss": 3.9356,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143247707131268e-05,
+      "loss": 3.9313,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9134861759620747e-05,
+      "loss": 3.945,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126475812110227e-05,
+      "loss": 3.9225,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9118106243403437e-05,
+      "loss": 3.9368,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9109736674696646e-05,
+      "loss": 3.941,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910135072718613e-05,
+      "loss": 3.9358,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.909296477967561e-05,
+      "loss": 3.9399,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.908457883216509e-05,
+      "loss": 3.9368,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.907619288465457e-05,
+      "loss": 3.9264,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067823315947778e-05,
+      "loss": 3.9395,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059437368437258e-05,
+      "loss": 3.9406,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9051051420926738e-05,
+      "loss": 3.9388,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904266547341622e-05,
+      "loss": 3.9439,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.90342795259057e-05,
+      "loss": 3.9444,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.902590995719891e-05,
+      "loss": 3.9413,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901752400968839e-05,
+      "loss": 3.9311,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900913806217787e-05,
+      "loss": 3.9374,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900075211466735e-05,
+      "loss": 3.9365,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.899236616715683e-05,
+      "loss": 3.9285,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8983980219646313e-05,
+      "loss": 3.9346,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8975594272135793e-05,
+      "loss": 3.9332,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8967208324625273e-05,
+      "loss": 3.9444,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958838755918482e-05,
+      "loss": 3.9286,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950452808407962e-05,
+      "loss": 3.9325,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8942066860897442e-05,
+      "loss": 3.9284,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933697292190655e-05,
+      "loss": 3.9363,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925311344680135e-05,
+      "loss": 3.931,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916925397169615e-05,
+      "loss": 3.9502,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908539449659095e-05,
+      "loss": 3.9384,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8900153502148575e-05,
+      "loss": 3.9333,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891767554638055e-05,
+      "loss": 3.9373,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883381607127535e-05,
+      "loss": 3.9386,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887499565961701e-05,
+      "loss": 3.944,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8866626090910227e-05,
+      "loss": 3.946,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8858240143399707e-05,
+      "loss": 3.9407,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849870574692916e-05,
+      "loss": 3.9476,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841484627182396e-05,
+      "loss": 3.9402,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8833098679671876e-05,
+      "loss": 3.9357,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8824712732161356e-05,
+      "loss": 3.932,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8816326784650832e-05,
+      "loss": 3.9385,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807940837140312e-05,
+      "loss": 3.9349,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8799554889629792e-05,
+      "loss": 3.9278,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791168942119272e-05,
+      "loss": 3.9354,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8782815752216217e-05,
+      "loss": 3.9307,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.87744298047057e-05,
+      "loss": 3.9396,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.876606023599891e-05,
+      "loss": 3.9375,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.875767428848839e-05,
+      "loss": 3.9356,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.008285045623779,
+      "eval_runtime": 289.0949,
+      "eval_samples_per_second": 1319.95,
+      "eval_steps_per_second": 41.249,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874928834097787e-05,
+      "loss": 3.9289,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874090239346735e-05,
+      "loss": 3.929,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.873251644595683e-05,
+      "loss": 3.9409,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8724130498446306e-05,
+      "loss": 3.9359,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715744550935786e-05,
+      "loss": 3.938,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707358603425266e-05,
+      "loss": 3.9316,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8698972655914746e-05,
+      "loss": 3.9255,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8690586708404226e-05,
+      "loss": 3.9305,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8682200760893706e-05,
+      "loss": 3.9423,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8673814813383186e-05,
+      "loss": 3.9384,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.86654452446764e-05,
+      "loss": 3.9353,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.865705929716588e-05,
+      "loss": 3.9449,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.864867334965536e-05,
+      "loss": 3.9203,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640287402144838e-05,
+      "loss": 3.9304,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631917833438047e-05,
+      "loss": 3.9307,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.862353188592753e-05,
+      "loss": 3.9224,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.861514593841701e-05,
+      "loss": 3.938,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.860675999090649e-05,
+      "loss": 3.9298,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.85983904221997e-05,
+      "loss": 3.929,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.859000447468918e-05,
+      "loss": 3.9593,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858161852717866e-05,
+      "loss": 3.9345,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.857323257966814e-05,
+      "loss": 3.9375,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564846632157623e-05,
+      "loss": 3.9381,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8556460684647103e-05,
+      "loss": 3.9489,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8548074737136583e-05,
+      "loss": 3.9303,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539688789626063e-05,
+      "loss": 3.9294,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531319220919272e-05,
+      "loss": 3.9335,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8522949652212484e-05,
+      "loss": 3.9324,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8514563704701964e-05,
+      "loss": 3.9318,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8506177757191444e-05,
+      "loss": 3.9282,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497791809680924e-05,
+      "loss": 3.9497,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489405862170404e-05,
+      "loss": 3.9381,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8481019914659884e-05,
+      "loss": 3.9364,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472633967149364e-05,
+      "loss": 3.9365,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464264398442577e-05,
+      "loss": 3.9384,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455878450932057e-05,
+      "loss": 3.936,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447492503421537e-05,
+      "loss": 3.9213,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8439106555911016e-05,
+      "loss": 3.9318,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.843072060840049e-05,
+      "loss": 3.9201,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.842233466088997e-05,
+      "loss": 3.9335,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8413948713379453e-05,
+      "loss": 3.9299,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8405562765868933e-05,
+      "loss": 3.9381,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8397176818358413e-05,
+      "loss": 3.9287,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8388823628455358e-05,
+      "loss": 3.9327,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8380437680944838e-05,
+      "loss": 3.9402,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8372051733434314e-05,
+      "loss": 3.9295,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8363665785923794e-05,
+      "loss": 3.937,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8355279838413274e-05,
+      "loss": 3.9227,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8346893890902754e-05,
+      "loss": 3.9282,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8338507943392234e-05,
+      "loss": 3.9364,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8330138374685443e-05,
+      "loss": 3.9327,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8321752427174923e-05,
+      "loss": 3.9373,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.831338285846814e-05,
+      "loss": 3.9108,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304996910957616e-05,
+      "loss": 3.936,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296610963447096e-05,
+      "loss": 3.9174,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8288225015936576e-05,
+      "loss": 3.946,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8279839068426056e-05,
+      "loss": 3.9177,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271453120915536e-05,
+      "loss": 3.9369,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8263067173405015e-05,
+      "loss": 3.9307,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8254681225894495e-05,
+      "loss": 3.9192,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.824629527838398e-05,
+      "loss": 3.9276,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.823790933087346e-05,
+      "loss": 3.9298,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.822952338336294e-05,
+      "loss": 3.9119,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.822113743585242e-05,
+      "loss": 3.9202,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212767867145628e-05,
+      "loss": 3.9364,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.820439829843884e-05,
+      "loss": 3.9278,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.819601235092832e-05,
+      "loss": 3.924,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81876264034178e-05,
+      "loss": 3.9215,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817924045590728e-05,
+      "loss": 3.9159,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817085450839676e-05,
+      "loss": 3.9306,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.816246856088624e-05,
+      "loss": 3.9332,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.815408261337572e-05,
+      "loss": 3.9336,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8145713044668932e-05,
+      "loss": 3.9403,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8137327097158412e-05,
+      "loss": 3.9338,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128941149647892e-05,
+      "loss": 3.9353,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8120555202137372e-05,
+      "loss": 3.9203,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8112169254626852e-05,
+      "loss": 3.9316,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8103783307116325e-05,
+      "loss": 3.9278,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.809539735960581e-05,
+      "loss": 3.9215,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.808701141209529e-05,
+      "loss": 3.9258,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8078641843388498e-05,
+      "loss": 3.9311,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070272274681714e-05,
+      "loss": 3.9412,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8061886327171194e-05,
+      "loss": 3.9327,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8053516758464403e-05,
+      "loss": 3.9167,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8045130810953886e-05,
+      "loss": 3.9235,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8036744863443366e-05,
+      "loss": 3.9386,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8028358915932846e-05,
+      "loss": 3.9343,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8019972968422326e-05,
+      "loss": 3.9305,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.80115870209118e-05,
+      "loss": 3.9396,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.800320107340128e-05,
+      "loss": 3.9351,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7994815125890762e-05,
+      "loss": 3.9199,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7986429178380242e-05,
+      "loss": 3.9294,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7978043230869722e-05,
+      "loss": 3.937,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7969657283359202e-05,
+      "loss": 3.9256,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7961271335848682e-05,
+      "loss": 3.9432,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.795290176714189e-05,
+      "loss": 3.9218,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.794451581963137e-05,
+      "loss": 3.923,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7936129872120855e-05,
+      "loss": 3.9278,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7927743924610335e-05,
+      "loss": 3.9356,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919357977099815e-05,
+      "loss": 3.9351,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7910972029589295e-05,
+      "loss": 3.9233,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7902586082078774e-05,
+      "loss": 3.9235,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7894200134568254e-05,
+      "loss": 3.9214,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885830565861464e-05,
+      "loss": 3.9368,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877444618350947e-05,
+      "loss": 3.9113,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7869058670840427e-05,
+      "loss": 3.9208,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7860689102133636e-05,
+      "loss": 3.934,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7852303154623116e-05,
+      "loss": 3.9272,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7843917207112596e-05,
+      "loss": 3.9241,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7835531259602076e-05,
+      "loss": 3.9271,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7827145312091556e-05,
+      "loss": 3.9201,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7818759364581036e-05,
+      "loss": 3.9303,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810389795874248e-05,
+      "loss": 3.9268,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7802003848363728e-05,
+      "loss": 3.9289,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7793617900853208e-05,
+      "loss": 3.9322,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7785231953342688e-05,
+      "loss": 3.9381,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7776846005832165e-05,
+      "loss": 3.9282,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768460058321645e-05,
+      "loss": 3.9212,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7760074110811125e-05,
+      "loss": 3.9295,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7751704542104334e-05,
+      "loss": 3.9271,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7743318594593814e-05,
+      "loss": 3.917,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734932647083294e-05,
+      "loss": 3.9245,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726546699572774e-05,
+      "loss": 3.9236,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7718160752062257e-05,
+      "loss": 3.9347,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7709774804551737e-05,
+      "loss": 3.9183,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7701388857041217e-05,
+      "loss": 3.9264,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7693002909530697e-05,
+      "loss": 3.912,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7684633340823906e-05,
+      "loss": 3.929,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7676247393313386e-05,
+      "loss": 3.9226,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667861445802866e-05,
+      "loss": 3.9387,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765947549829235e-05,
+      "loss": 3.9297,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7651105929585558e-05,
+      "loss": 3.9289,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7642719982075038e-05,
+      "loss": 3.9219,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634334034564518e-05,
+      "loss": 3.9269,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7625948087053998e-05,
+      "loss": 3.9368,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.761757851834721e-05,
+      "loss": 3.9395,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.760919257083669e-05,
+      "loss": 3.9316,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.760080662332617e-05,
+      "loss": 3.937,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.759242067581565e-05,
+      "loss": 3.9276,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.758405110710886e-05,
+      "loss": 3.9266,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.757566515959834e-05,
+      "loss": 3.9239,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.756727921208782e-05,
+      "loss": 3.9281,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7558893264577303e-05,
+      "loss": 3.9274,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7550523695870512e-05,
+      "loss": 3.9175,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7542137748359992e-05,
+      "loss": 3.9273,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.75337681796532e-05,
+      "loss": 3.9227,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.752538223214268e-05,
+      "loss": 3.9308,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7516996284632164e-05,
+      "loss": 3.9263,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508610337121644e-05,
+      "loss": 3.9268,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.003251552581787,
+      "eval_runtime": 314.9348,
+      "eval_samples_per_second": 1211.651,
+      "eval_steps_per_second": 37.865,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7500224389611124e-05,
+      "loss": 3.9163,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491838442100604e-05,
+      "loss": 3.9213,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483452494590084e-05,
+      "loss": 3.9291,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7475066547079564e-05,
+      "loss": 3.9267,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466680599569044e-05,
+      "loss": 3.9328,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.745829465205852e-05,
+      "loss": 3.9166,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7449908704548e-05,
+      "loss": 3.9215,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744152275703748e-05,
+      "loss": 3.9194,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743313680952696e-05,
+      "loss": 3.9314,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742475086201644e-05,
+      "loss": 3.9287,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741636491450592e-05,
+      "loss": 3.9277,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.74079789669954e-05,
+      "loss": 3.9303,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739959301948488e-05,
+      "loss": 3.9122,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7391207071974364e-05,
+      "loss": 3.9165,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7382821124463843e-05,
+      "loss": 3.9231,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374435176953323e-05,
+      "loss": 3.9151,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7366049229442803e-05,
+      "loss": 3.9245,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7357663281932283e-05,
+      "loss": 3.9207,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7349277334421763e-05,
+      "loss": 3.9201,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7340907765714972e-05,
+      "loss": 3.9501,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332521818204452e-05,
+      "loss": 3.9228,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7324135870693936e-05,
+      "loss": 3.9311,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.731574992318341e-05,
+      "loss": 3.9268,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.730736397567289e-05,
+      "loss": 3.9385,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.729897802816237e-05,
+      "loss": 3.9184,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.729059208065185e-05,
+      "loss": 3.9214,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.728220613314133e-05,
+      "loss": 3.9235,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.727383656443454e-05,
+      "loss": 3.9245,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.726545061692402e-05,
+      "loss": 3.9169,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.72570646694135e-05,
+      "loss": 3.9231,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724867872190298e-05,
+      "loss": 3.9368,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724030915319619e-05,
+      "loss": 3.929,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7231939584489406e-05,
+      "loss": 3.9241,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223553636978883e-05,
+      "loss": 3.9319,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.72151840682721e-05,
+      "loss": 3.9268,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.720679812076158e-05,
+      "loss": 3.9251,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.719841217325106e-05,
+      "loss": 3.9094,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7190026225740535e-05,
+      "loss": 3.9258,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7181640278230015e-05,
+      "loss": 3.9124,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7173254330719495e-05,
+      "loss": 3.9222,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7164868383208975e-05,
+      "loss": 3.9196,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156482435698455e-05,
+      "loss": 3.929,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7148096488187935e-05,
+      "loss": 3.9215,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139726919481144e-05,
+      "loss": 3.9206,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131340971970627e-05,
+      "loss": 3.929,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122955024460107e-05,
+      "loss": 3.9227,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7114569076949587e-05,
+      "loss": 3.9287,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7106183129439067e-05,
+      "loss": 3.9142,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7097797181928547e-05,
+      "loss": 3.9156,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089411234418027e-05,
+      "loss": 3.9287,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7081025286907507e-05,
+      "loss": 3.9225,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.707265571820072e-05,
+      "loss": 3.9249,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.70642697706902e-05,
+      "loss": 3.9055,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.705588382317968e-05,
+      "loss": 3.9287,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.704749787566916e-05,
+      "loss": 3.9089,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703912830696237e-05,
+      "loss": 3.9352,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7030742359451848e-05,
+      "loss": 3.9051,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7022356411941328e-05,
+      "loss": 3.9294,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.701397046443081e-05,
+      "loss": 3.9185,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.700558451692029e-05,
+      "loss": 3.9148,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.69972149482135e-05,
+      "loss": 3.9157,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698882900070298e-05,
+      "loss": 3.9194,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698044305319246e-05,
+      "loss": 3.9024,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.697205710568194e-05,
+      "loss": 3.912,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6963687536975153e-05,
+      "loss": 3.9212,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6955301589464633e-05,
+      "loss": 3.9224,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6946915641954113e-05,
+      "loss": 3.9119,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6938546073247322e-05,
+      "loss": 3.9129,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6930160125736802e-05,
+      "loss": 3.9131,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921774178226282e-05,
+      "loss": 3.9161,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913388230715765e-05,
+      "loss": 3.9216,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6905018662008974e-05,
+      "loss": 3.9223,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6896632714498454e-05,
+      "loss": 3.9297,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6888246766987934e-05,
+      "loss": 3.929,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6879860819477414e-05,
+      "loss": 3.9249,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871474871966894e-05,
+      "loss": 3.9088,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.686308892445637e-05,
+      "loss": 3.9199,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.685470297694585e-05,
+      "loss": 3.922,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.684631702943533e-05,
+      "loss": 3.908,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.683794746072854e-05,
+      "loss": 3.9197,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682956151321802e-05,
+      "loss": 3.9217,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6821175565707503e-05,
+      "loss": 3.9297,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6812789618196983e-05,
+      "loss": 3.9197,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804420049490192e-05,
+      "loss": 3.9044,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6796034101979672e-05,
+      "loss": 3.9181,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787648154469152e-05,
+      "loss": 3.9241,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779262206958632e-05,
+      "loss": 3.9255,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770892638251844e-05,
+      "loss": 3.92,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6762506690741324e-05,
+      "loss": 3.9301,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6754120743230804e-05,
+      "loss": 3.9257,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745751174524013e-05,
+      "loss": 3.9078,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737365227013493e-05,
+      "loss": 3.9225,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728979279502973e-05,
+      "loss": 3.9234,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720593331992457e-05,
+      "loss": 3.9136,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6712207384481937e-05,
+      "loss": 3.9364,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703837815775146e-05,
+      "loss": 3.9184,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695451868264626e-05,
+      "loss": 3.9097,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6687065920754106e-05,
+      "loss": 3.9196,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6678679973243586e-05,
+      "loss": 3.9284,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670294025733066e-05,
+      "loss": 3.9232,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6661924457026278e-05,
+      "loss": 3.9171,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653538509515758e-05,
+      "loss": 3.9123,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6645152562005238e-05,
+      "loss": 3.9115,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6636766614494718e-05,
+      "loss": 3.9295,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6628380666984198e-05,
+      "loss": 3.9018,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6619994719473678e-05,
+      "loss": 3.9151,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6611608771963158e-05,
+      "loss": 3.9244,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6603222824452638e-05,
+      "loss": 3.916,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.659485325574585e-05,
+      "loss": 3.9159,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.658646730823533e-05,
+      "loss": 3.9192,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657808136072481e-05,
+      "loss": 3.9152,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656969541321429e-05,
+      "loss": 3.9207,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656130946570377e-05,
+      "loss": 3.9174,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6552939896996983e-05,
+      "loss": 3.9165,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6544553949486463e-05,
+      "loss": 3.9251,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6536168001975942e-05,
+      "loss": 3.9289,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.652779843326915e-05,
+      "loss": 3.9239,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.651941248575863e-05,
+      "loss": 3.9086,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.651102653824811e-05,
+      "loss": 3.917,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.650264059073759e-05,
+      "loss": 3.9162,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6494271022030804e-05,
+      "loss": 3.9067,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6485885074520284e-05,
+      "loss": 3.9206,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6477499127009764e-05,
+      "loss": 3.9086,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6469113179499244e-05,
+      "loss": 3.9259,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460727231988724e-05,
+      "loss": 3.9084,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452341284478204e-05,
+      "loss": 3.9163,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.644395533696768e-05,
+      "loss": 3.9051,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6435585768260896e-05,
+      "loss": 3.9195,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6427199820750376e-05,
+      "loss": 3.9101,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641881387323985e-05,
+      "loss": 3.9299,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641042792572933e-05,
+      "loss": 3.9258,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6402041978218813e-05,
+      "loss": 3.9178,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6393656030708293e-05,
+      "loss": 3.9104,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385270083197772e-05,
+      "loss": 3.9171,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6376884135687252e-05,
+      "loss": 3.9266,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368498188176732e-05,
+      "loss": 3.931,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6360112240666212e-05,
+      "loss": 3.9201,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6351726293155692e-05,
+      "loss": 3.9268,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343340345645172e-05,
+      "loss": 3.9188,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334970776938385e-05,
+      "loss": 3.9183,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6326601208231594e-05,
+      "loss": 3.9134,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6318215260721074e-05,
+      "loss": 3.915,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309829313210554e-05,
+      "loss": 3.9198,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301443365700034e-05,
+      "loss": 3.9105,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6293057418189514e-05,
+      "loss": 3.9147,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284671470678997e-05,
+      "loss": 3.9187,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276285523168477e-05,
+      "loss": 3.9136,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6267915954461686e-05,
+      "loss": 3.9162,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259530006951166e-05,
+      "loss": 3.9145,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9987401962280273,
+      "eval_runtime": 312.2299,
+      "eval_samples_per_second": 1222.147,
+      "eval_steps_per_second": 38.193,
+      "step": 1450080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6251144059440646e-05,
+      "loss": 3.9147,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6242758111930126e-05,
+      "loss": 3.9093,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6234372164419606e-05,
+      "loss": 3.9237,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6225986216909086e-05,
+      "loss": 3.9159,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6217600269398562e-05,
+      "loss": 3.9269,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6209214321888042e-05,
+      "loss": 3.9065,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6200844753181258e-05,
+      "loss": 3.9139,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6192458805670738e-05,
+      "loss": 3.9075,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6184072858160215e-05,
+      "loss": 3.9236,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6175686910649695e-05,
+      "loss": 3.9194,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.616731734194291e-05,
+      "loss": 3.9193,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.615893139443239e-05,
+      "loss": 3.9202,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.61505618257256e-05,
+      "loss": 3.9026,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.614217587821508e-05,
+      "loss": 3.9094,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.613378993070456e-05,
+      "loss": 3.9132,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6125403983194036e-05,
+      "loss": 3.9097,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6117018035683516e-05,
+      "loss": 3.9128,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6108632088172996e-05,
+      "loss": 3.9164,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6100262519466212e-05,
+      "loss": 3.9069,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.609187657195569e-05,
+      "loss": 3.9442,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.608349062444517e-05,
+      "loss": 3.9138,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6075121055738384e-05,
+      "loss": 3.9199,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6066735108227864e-05,
+      "loss": 3.9199,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6058349160717337e-05,
+      "loss": 3.9274,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6049963213206817e-05,
+      "loss": 3.9099,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6041577265696297e-05,
+      "loss": 3.9187,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6033191318185777e-05,
+      "loss": 3.9127,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.602480537067526e-05,
+      "loss": 3.9158,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.601641942316474e-05,
+      "loss": 3.9067,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.600803347565422e-05,
+      "loss": 3.9146,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599966390694743e-05,
+      "loss": 3.9261,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599127795943691e-05,
+      "loss": 3.9193,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.598289201192639e-05,
+      "loss": 3.9209,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.597450606441587e-05,
+      "loss": 3.9203,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5966136495709082e-05,
+      "loss": 3.9135,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957750548198562e-05,
+      "loss": 3.9214,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5949364600688042e-05,
+      "loss": 3.9009,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5940978653177522e-05,
+      "loss": 3.9167,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.593260908447073e-05,
+      "loss": 3.9022,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5924223136960214e-05,
+      "loss": 3.9154,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5915837189449694e-05,
+      "loss": 3.9113,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5907451241939174e-05,
+      "loss": 3.9206,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5899065294428654e-05,
+      "loss": 3.9137,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5890695725721863e-05,
+      "loss": 3.9113,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5882309778211343e-05,
+      "loss": 3.9204,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5873923830700823e-05,
+      "loss": 3.9162,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5865537883190307e-05,
+      "loss": 3.9172,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5857151935679786e-05,
+      "loss": 3.9099,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5848765988169266e-05,
+      "loss": 3.9031,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5840380040658746e-05,
+      "loss": 3.9198,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5831994093148226e-05,
+      "loss": 3.9132,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5823624524441435e-05,
+      "loss": 3.9172,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5815238576930915e-05,
+      "loss": 3.9014,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.58068526294204e-05,
+      "loss": 3.9222,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5798483060713608e-05,
+      "loss": 3.8958,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5790097113203088e-05,
+      "loss": 3.9286,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781711165692568e-05,
+      "loss": 3.8997,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5773325218182048e-05,
+      "loss": 3.9152,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5764939270671524e-05,
+      "loss": 3.9107,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5756553323161004e-05,
+      "loss": 3.9097,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5748167375650484e-05,
+      "loss": 3.9059,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5739781428139964e-05,
+      "loss": 3.9172,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731411859433173e-05,
+      "loss": 3.8933,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5723025911922653e-05,
+      "loss": 3.8994,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5714639964412137e-05,
+      "loss": 3.9194,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5706270395705346e-05,
+      "loss": 3.9118,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5697884448194826e-05,
+      "loss": 3.9068,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5689498500684306e-05,
+      "loss": 3.9073,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5681112553173786e-05,
+      "loss": 3.8968,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5672726605663265e-05,
+      "loss": 3.9078,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5664340658152745e-05,
+      "loss": 3.9154,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.565595471064223e-05,
+      "loss": 3.915,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.564756876313171e-05,
+      "loss": 3.9218,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5639199194424918e-05,
+      "loss": 3.9184,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5630813246914398e-05,
+      "loss": 3.9191,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5622443678207607e-05,
+      "loss": 3.9021,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.561405773069709e-05,
+      "loss": 3.91,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.560567178318657e-05,
+      "loss": 3.9172,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.559728583567605e-05,
+      "loss": 3.9,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.558889988816553e-05,
+      "loss": 3.9161,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.558051394065501e-05,
+      "loss": 3.9109,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.557212799314449e-05,
+      "loss": 3.9191,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.556374204563397e-05,
+      "loss": 3.9149,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.555538885573091e-05,
+      "loss": 3.8992,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.554700290822039e-05,
+      "loss": 3.9117,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553861696070987e-05,
+      "loss": 3.9137,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553023101319935e-05,
+      "loss": 3.918,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.552186144449256e-05,
+      "loss": 3.9135,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5513475496982044e-05,
+      "loss": 3.919,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5505089549471524e-05,
+      "loss": 3.9217,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5496703601961004e-05,
+      "loss": 3.8999,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5488317654450484e-05,
+      "loss": 3.9164,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5479931706939964e-05,
+      "loss": 3.9101,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5471545759429444e-05,
+      "loss": 3.9085,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5463159811918924e-05,
+      "loss": 3.9253,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5454773864408404e-05,
+      "loss": 3.9121,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5446404295701616e-05,
+      "loss": 3.9002,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5438018348191096e-05,
+      "loss": 3.9112,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5429632400680576e-05,
+      "loss": 3.9214,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5421262831973785e-05,
+      "loss": 3.9164,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5412876884463265e-05,
+      "loss": 3.9084,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5404490936952745e-05,
+      "loss": 3.9014,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5396104989442225e-05,
+      "loss": 3.9047,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5387719041931708e-05,
+      "loss": 3.9215,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5379349473224917e-05,
+      "loss": 3.8942,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5370963525714397e-05,
+      "loss": 3.9061,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5362577578203877e-05,
+      "loss": 3.9149,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5354208009497086e-05,
+      "loss": 3.9109,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.534582206198657e-05,
+      "loss": 3.9037,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.533743611447605e-05,
+      "loss": 3.9133,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.532905016696553e-05,
+      "loss": 3.9086,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5320664219455003e-05,
+      "loss": 3.9101,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5312278271944483e-05,
+      "loss": 3.9107,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5303892324433963e-05,
+      "loss": 3.9093,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5295506376923446e-05,
+      "loss": 3.9222,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5287120429412926e-05,
+      "loss": 3.9157,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.527876723950987e-05,
+      "loss": 3.9172,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.527038129199935e-05,
+      "loss": 3.9009,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.526199534448883e-05,
+      "loss": 3.9105,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5253609396978308e-05,
+      "loss": 3.9111,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5245223449467787e-05,
+      "loss": 3.9001,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5236837501957267e-05,
+      "loss": 3.913,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5228451554446747e-05,
+      "loss": 3.9048,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5220065606936227e-05,
+      "loss": 3.9151,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5211696038229436e-05,
+      "loss": 3.9024,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5203310090718916e-05,
+      "loss": 3.9102,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.51949241432084e-05,
+      "loss": 3.893,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.518655457450161e-05,
+      "loss": 3.913,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.517816862699109e-05,
+      "loss": 3.9043,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.516978267948057e-05,
+      "loss": 3.9196,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.516139673197005e-05,
+      "loss": 3.9165,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.515301078445953e-05,
+      "loss": 3.9163,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.514464121575274e-05,
+      "loss": 3.9015,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.513627164704595e-05,
+      "loss": 3.9129,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.512788569953543e-05,
+      "loss": 3.9159,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511949975202491e-05,
+      "loss": 3.9236,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511111380451439e-05,
+      "loss": 3.914,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.510272785700387e-05,
+      "loss": 3.9155,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5094341909493353e-05,
+      "loss": 3.9139,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5085955961982833e-05,
+      "loss": 3.9155,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5077570014472313e-05,
+      "loss": 3.9028,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5069184066961793e-05,
+      "loss": 3.908,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5060798119451273e-05,
+      "loss": 3.9095,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5052428550744482e-05,
+      "loss": 3.9051,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5044042603233962e-05,
+      "loss": 3.9065,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5035656655723446e-05,
+      "loss": 3.9136,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5027270708212926e-05,
+      "loss": 3.9062,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5018901139506135e-05,
+      "loss": 3.9073,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5010515191995615e-05,
+      "loss": 3.9159,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.996168851852417,
+      "eval_runtime": 293.7871,
+      "eval_samples_per_second": 1298.869,
+      "eval_steps_per_second": 40.591,
+      "step": 1526400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5002129244485095e-05,
+      "loss": 3.8997,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.499374329697457e-05,
+      "loss": 3.9054,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4985357349464054e-05,
+      "loss": 3.9148,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4976971401953534e-05,
+      "loss": 3.908,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4968585454443014e-05,
+      "loss": 3.9218,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4960199506932494e-05,
+      "loss": 3.8981,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4951813559421974e-05,
+      "loss": 3.9044,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4943427611911454e-05,
+      "loss": 3.8993,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.493504166440093e-05,
+      "loss": 3.9164,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.492665571689041e-05,
+      "loss": 3.9156,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4918269769379894e-05,
+      "loss": 3.9044,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4909883821869374e-05,
+      "loss": 3.9199,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4901497874358854e-05,
+      "loss": 3.893,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4893111926848334e-05,
+      "loss": 3.9026,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4884725979337814e-05,
+      "loss": 3.9103,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.487634003182729e-05,
+      "loss": 3.9009,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.486795408431677e-05,
+      "loss": 3.9068,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4859584515609986e-05,
+      "loss": 3.9066,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4851198568099463e-05,
+      "loss": 3.9049,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4842812620588943e-05,
+      "loss": 3.9332,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4834426673078423e-05,
+      "loss": 3.9076,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4826040725567903e-05,
+      "loss": 3.9095,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4817654778057383e-05,
+      "loss": 3.9147,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4809268830546863e-05,
+      "loss": 3.9191,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4800882883036343e-05,
+      "loss": 3.9007,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4792513314329555e-05,
+      "loss": 3.9143,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4784127366819035e-05,
+      "loss": 3.9023,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4775757798112244e-05,
+      "loss": 3.9084,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4767371850601724e-05,
+      "loss": 3.8997,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4758985903091207e-05,
+      "loss": 3.9083,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4750599955580687e-05,
+      "loss": 3.9212,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4742214008070167e-05,
+      "loss": 3.9128,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4733828060559644e-05,
+      "loss": 3.9125,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4725442113049124e-05,
+      "loss": 3.9129,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4717072544342336e-05,
+      "loss": 3.9077,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4708686596831816e-05,
+      "loss": 3.9136,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4700300649321296e-05,
+      "loss": 3.8924,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4691914701810776e-05,
+      "loss": 3.9092,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4683528754300256e-05,
+      "loss": 3.9001,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4675142806789736e-05,
+      "loss": 3.9019,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4666756859279216e-05,
+      "loss": 3.9067,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4658370911768696e-05,
+      "loss": 3.9134,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.465000134306191e-05,
+      "loss": 3.9064,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464161539555139e-05,
+      "loss": 3.9074,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.463322944804087e-05,
+      "loss": 3.9129,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.462484350053035e-05,
+      "loss": 3.9053,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4616473931823558e-05,
+      "loss": 3.9109,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4608087984313037e-05,
+      "loss": 3.9057,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4599702036802517e-05,
+      "loss": 3.8939,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4591316089292e-05,
+      "loss": 3.9142,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.458294652058521e-05,
+      "loss": 3.903,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.457456057307469e-05,
+      "loss": 3.9097,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.456617462556417e-05,
+      "loss": 3.897,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557788678053646e-05,
+      "loss": 3.9115,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549402730543126e-05,
+      "loss": 3.891,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4541033161836342e-05,
+      "loss": 3.9187,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4532647214325822e-05,
+      "loss": 3.8912,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.45242612668153e-05,
+      "loss": 3.9026,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.451587531930478e-05,
+      "loss": 3.9072,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.450748937179426e-05,
+      "loss": 3.9013,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.449911980308747e-05,
+      "loss": 3.9019,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4490750234380684e-05,
+      "loss": 3.9054,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4482364286870164e-05,
+      "loss": 3.8928,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4473978339359644e-05,
+      "loss": 3.8906,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.446559239184912e-05,
+      "loss": 3.9079,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.44572064443386e-05,
+      "loss": 3.9071,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4448836875631816e-05,
+      "loss": 3.8965,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4440450928121296e-05,
+      "loss": 3.904,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4432064980610772e-05,
+      "loss": 3.8894,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4423695411903985e-05,
+      "loss": 3.902,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4415309464393465e-05,
+      "loss": 3.9079,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4406923516882945e-05,
+      "loss": 3.9042,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4398537569372425e-05,
+      "loss": 3.914,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4390151621861905e-05,
+      "loss": 3.9116,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4381782053155117e-05,
+      "loss": 3.9148,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373396105644594e-05,
+      "loss": 3.8943,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4365010158134074e-05,
+      "loss": 3.9018,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4356624210623554e-05,
+      "loss": 3.9077,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4348238263113034e-05,
+      "loss": 3.8875,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4339852315602517e-05,
+      "loss": 3.9115,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4331466368091997e-05,
+      "loss": 3.9042,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4323080420581477e-05,
+      "loss": 3.9119,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4314694473070953e-05,
+      "loss": 3.9089,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4306308525560433e-05,
+      "loss": 3.8909,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4297922578049913e-05,
+      "loss": 3.9059,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4289536630539393e-05,
+      "loss": 3.9053,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4281167061832606e-05,
+      "loss": 3.9103,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4272781114322086e-05,
+      "loss": 3.9049,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4264395166811566e-05,
+      "loss": 3.914,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4256025598104775e-05,
+      "loss": 3.9119,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4247639650594255e-05,
+      "loss": 3.8955,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4239253703083735e-05,
+      "loss": 3.9085,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4230867755573218e-05,
+      "loss": 3.8997,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4222481808062698e-05,
+      "loss": 3.9087,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4214095860552178e-05,
+      "loss": 3.9188,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205709913041658e-05,
+      "loss": 3.9012,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4197323965531135e-05,
+      "loss": 3.8967,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.418897077562808e-05,
+      "loss": 3.9037,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.418058482811756e-05,
+      "loss": 3.9091,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.417219888060704e-05,
+      "loss": 3.9146,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.416381293309652e-05,
+      "loss": 3.8985,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4155426985586e-05,
+      "loss": 3.8978,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.414704103807548e-05,
+      "loss": 3.896,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4138655090564956e-05,
+      "loss": 3.9138,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4130269143054436e-05,
+      "loss": 3.8887,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.412191595315138e-05,
+      "loss": 3.8979,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.411353000564086e-05,
+      "loss": 3.9073,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.410514405813034e-05,
+      "loss": 3.9017,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.409675811061982e-05,
+      "loss": 3.8996,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.40883721631093e-05,
+      "loss": 3.9089,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.407998621559878e-05,
+      "loss": 3.9009,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.407160026808826e-05,
+      "loss": 3.8998,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.406321432057774e-05,
+      "loss": 3.9055,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4054844751870953e-05,
+      "loss": 3.9052,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.404645880436043e-05,
+      "loss": 3.9115,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.403807285684991e-05,
+      "loss": 3.9094,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4029719666946855e-05,
+      "loss": 3.9106,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4021333719436335e-05,
+      "loss": 3.8938,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4012947771925814e-05,
+      "loss": 3.9038,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4004561824415294e-05,
+      "loss": 3.9041,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3996175876904774e-05,
+      "loss": 3.8914,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3987789929394254e-05,
+      "loss": 3.9099,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3979403981883734e-05,
+      "loss": 3.8955,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3971018034373214e-05,
+      "loss": 3.9081,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3962632086862694e-05,
+      "loss": 3.902,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3954246139352174e-05,
+      "loss": 3.9042,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3945860191841654e-05,
+      "loss": 3.8849,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3937474244331134e-05,
+      "loss": 3.9024,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3929104675624343e-05,
+      "loss": 3.9011,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3920718728113827e-05,
+      "loss": 3.9055,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3912332780603306e-05,
+      "loss": 3.9095,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3903946833092786e-05,
+      "loss": 3.9121,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3895577264385996e-05,
+      "loss": 3.8903,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3887207695679205e-05,
+      "loss": 3.9069,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3878821748168688e-05,
+      "loss": 3.9122,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3870435800658168e-05,
+      "loss": 3.9129,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3862049853147648e-05,
+      "loss": 3.9081,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3853680284440857e-05,
+      "loss": 3.912,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.384531071573407e-05,
+      "loss": 3.9052,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.383692476822355e-05,
+      "loss": 3.9088,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382853882071303e-05,
+      "loss": 3.8941,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382015287320251e-05,
+      "loss": 3.9043,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381176692569199e-05,
+      "loss": 3.9002,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.380338097818147e-05,
+      "loss": 3.9007,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.379499503067095e-05,
+      "loss": 3.9007,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.378660908316043e-05,
+      "loss": 3.9008,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.377822313564991e-05,
+      "loss": 3.9011,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.376985356694312e-05,
+      "loss": 3.8987,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.37614676194326e-05,
+      "loss": 3.9097,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.992966413497925,
+      "eval_runtime": 298.5244,
+      "eval_samples_per_second": 1278.257,
+      "eval_steps_per_second": 39.946,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.375308167192208e-05,
+      "loss": 3.8966,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3744695724411558e-05,
+      "loss": 3.8949,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3736309776901038e-05,
+      "loss": 3.9063,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3727923829390518e-05,
+      "loss": 3.906,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3719537881879998e-05,
+      "loss": 3.9077,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.371115193436948e-05,
+      "loss": 3.8979,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.370276598685896e-05,
+      "loss": 3.8947,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.369438003934844e-05,
+      "loss": 3.897,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3685994091837918e-05,
+      "loss": 3.9015,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3677608144327398e-05,
+      "loss": 3.9108,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3669222196816878e-05,
+      "loss": 3.8976,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3660836249306358e-05,
+      "loss": 3.9099,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.365245030179584e-05,
+      "loss": 3.8906,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.364406435428532e-05,
+      "loss": 3.8932,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3635678406774797e-05,
+      "loss": 3.9027,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3627292459264277e-05,
+      "loss": 3.8901,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3618906511753757e-05,
+      "loss": 3.8984,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3610520564243237e-05,
+      "loss": 3.9005,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.360215099553645e-05,
+      "loss": 3.8952,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.359376504802593e-05,
+      "loss": 3.9238,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.358537910051541e-05,
+      "loss": 3.9055,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.357699315300489e-05,
+      "loss": 3.9034,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356860720549437e-05,
+      "loss": 3.9073,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356023763678758e-05,
+      "loss": 3.9145,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.355185168927706e-05,
+      "loss": 3.8942,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3543465741766542e-05,
+      "loss": 3.9042,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3535079794256022e-05,
+      "loss": 3.8993,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3526693846745502e-05,
+      "loss": 3.898,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.351830789923498e-05,
+      "loss": 3.8897,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.350992195172446e-05,
+      "loss": 3.9034,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.350153600421394e-05,
+      "loss": 3.9141,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.349316643550715e-05,
+      "loss": 3.905,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.348478048799663e-05,
+      "loss": 3.9085,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.347639454048611e-05,
+      "loss": 3.9031,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3468024971779323e-05,
+      "loss": 3.9025,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3459639024268803e-05,
+      "loss": 3.9059,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.345125307675828e-05,
+      "loss": 3.8929,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.344286712924776e-05,
+      "loss": 3.8977,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3434481181737243e-05,
+      "loss": 3.8936,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3426095234226723e-05,
+      "loss": 3.8943,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3417709286716203e-05,
+      "loss": 3.9018,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3409323339205683e-05,
+      "loss": 3.9069,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.340093739169516e-05,
+      "loss": 3.8961,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.339255144418464e-05,
+      "loss": 3.9,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.338416549667412e-05,
+      "loss": 3.9081,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.33757795491636e-05,
+      "loss": 3.8993,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3367409980456812e-05,
+      "loss": 3.9027,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3359024032946292e-05,
+      "loss": 3.8968,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3350638085435772e-05,
+      "loss": 3.8866,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3342252137925252e-05,
+      "loss": 3.9082,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3333866190414732e-05,
+      "loss": 3.8959,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3325480242904212e-05,
+      "loss": 3.9046,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.331709429539369e-05,
+      "loss": 3.8902,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.330870834788317e-05,
+      "loss": 3.9016,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3300338779176384e-05,
+      "loss": 3.8843,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3291969210469593e-05,
+      "loss": 3.9084,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3283583262959073e-05,
+      "loss": 3.8872,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3275197315448553e-05,
+      "loss": 3.8984,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3266811367938036e-05,
+      "loss": 3.9012,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3258425420427516e-05,
+      "loss": 3.8906,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3250039472916993e-05,
+      "loss": 3.8911,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3241653525406473e-05,
+      "loss": 3.9051,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3233267577895953e-05,
+      "loss": 3.883,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3224898009189165e-05,
+      "loss": 3.8812,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3216512061678645e-05,
+      "loss": 3.9001,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3208126114168125e-05,
+      "loss": 3.9032,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3199740166657605e-05,
+      "loss": 3.8894,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3191370597950814e-05,
+      "loss": 3.9,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3182984650440294e-05,
+      "loss": 3.8795,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3174598702929774e-05,
+      "loss": 3.8925,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3166212755419258e-05,
+      "loss": 3.9017,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3157826807908738e-05,
+      "loss": 3.9005,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3149440860398218e-05,
+      "loss": 3.9024,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3141054912887697e-05,
+      "loss": 3.9052,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3132668965377174e-05,
+      "loss": 3.9111,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.312431577547412e-05,
+      "loss": 3.8853,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.31159298279636e-05,
+      "loss": 3.899,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.310754388045308e-05,
+      "loss": 3.9037,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309915793294256e-05,
+      "loss": 3.8789,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309077198543204e-05,
+      "loss": 3.9021,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.308238603792152e-05,
+      "loss": 3.9001,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3074000090410995e-05,
+      "loss": 3.903,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3065614142900475e-05,
+      "loss": 3.903,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.305724457419369e-05,
+      "loss": 3.8891,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.304885862668317e-05,
+      "loss": 3.8932,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.304048905797638e-05,
+      "loss": 3.8972,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.303210311046586e-05,
+      "loss": 3.9065,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.302371716295534e-05,
+      "loss": 3.8946,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.301533121544482e-05,
+      "loss": 3.9087,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.30069452679343e-05,
+      "loss": 3.9073,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.299855932042378e-05,
+      "loss": 3.8897,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.299017337291326e-05,
+      "loss": 3.9007,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.298180380420647e-05,
+      "loss": 3.8914,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.297341785669595e-05,
+      "loss": 3.9024,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.296503190918543e-05,
+      "loss": 3.914,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2956645961674912e-05,
+      "loss": 3.8912,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2948260014164392e-05,
+      "loss": 3.8936,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2939874066653872e-05,
+      "loss": 3.8902,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2931488119143352e-05,
+      "loss": 3.9029,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.292310217163283e-05,
+      "loss": 3.9107,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.291473260292604e-05,
+      "loss": 3.8929,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.290634665541552e-05,
+      "loss": 3.8921,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2897960707905e-05,
+      "loss": 3.8851,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2889591139198214e-05,
+      "loss": 3.9069,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2881205191687694e-05,
+      "loss": 3.879,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2872819244177174e-05,
+      "loss": 3.8957,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.286443329666665e-05,
+      "loss": 3.8959,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.285604734915613e-05,
+      "loss": 3.8977,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2847661401645613e-05,
+      "loss": 3.8944,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2839275454135093e-05,
+      "loss": 3.8997,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2830889506624573e-05,
+      "loss": 3.8955,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2822536316721515e-05,
+      "loss": 3.892,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2814166748014727e-05,
+      "loss": 3.8976,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2805780800504207e-05,
+      "loss": 3.8984,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2797394852993687e-05,
+      "loss": 3.9027,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2789008905483167e-05,
+      "loss": 3.907,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2780622957972647e-05,
+      "loss": 3.8989,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2772237010462124e-05,
+      "loss": 3.8875,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2763851062951604e-05,
+      "loss": 3.8961,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.275548149424482e-05,
+      "loss": 3.8982,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2747095546734296e-05,
+      "loss": 3.8874,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2738709599223776e-05,
+      "loss": 3.8985,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2730323651713256e-05,
+      "loss": 3.8876,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2721937704202736e-05,
+      "loss": 3.8996,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2713551756692216e-05,
+      "loss": 3.8945,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2705165809181696e-05,
+      "loss": 3.8935,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2696779861671176e-05,
+      "loss": 3.8867,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2688393914160656e-05,
+      "loss": 3.8922,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268002434545387e-05,
+      "loss": 3.8951,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.267163839794335e-05,
+      "loss": 3.9034,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.266325245043283e-05,
+      "loss": 3.8987,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2654866502922305e-05,
+      "loss": 3.9036,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2646480555411785e-05,
+      "loss": 3.8888,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2638110986705e-05,
+      "loss": 3.898,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2629725039194477e-05,
+      "loss": 3.9063,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2621339091683957e-05,
+      "loss": 3.9046,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2612953144173437e-05,
+      "loss": 3.9008,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2604567196662917e-05,
+      "loss": 3.907,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2596181249152397e-05,
+      "loss": 3.8962,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2587795301641877e-05,
+      "loss": 3.9036,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257940935413136e-05,
+      "loss": 3.8869,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257103978542457e-05,
+      "loss": 3.8962,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.256265383791405e-05,
+      "loss": 3.893,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.255426789040353e-05,
+      "loss": 3.8936,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.254588194289301e-05,
+      "loss": 3.8943,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2537512374186222e-05,
+      "loss": 3.8918,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2529126426675702e-05,
+      "loss": 3.9016,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.252075685796891e-05,
+      "loss": 3.8835,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.251237091045839e-05,
+      "loss": 3.9062,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.989954710006714,
+      "eval_runtime": 297.5717,
+      "eval_samples_per_second": 1282.35,
+      "eval_steps_per_second": 40.074,
+      "step": 1679040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.250398496294787e-05,
+      "loss": 3.8902,
+      "step": 1679360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.249559901543735e-05,
+      "loss": 3.8896,
+      "step": 1679872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.248721306792683e-05,
+      "loss": 3.8986,
+      "step": 1680384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.247882712041631e-05,
+      "loss": 3.8988,
+      "step": 1680896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2470457551709523e-05,
+      "loss": 3.9003,
+      "step": 1681408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2462071604199003e-05,
+      "loss": 3.891,
+      "step": 1681920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2453685656688483e-05,
+      "loss": 3.8841,
+      "step": 1682432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.244529970917796e-05,
+      "loss": 3.893,
+      "step": 1682944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.243691376166744e-05,
+      "loss": 3.8968,
+      "step": 1683456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2428527814156923e-05,
+      "loss": 3.9038,
+      "step": 1683968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2420141866646403e-05,
+      "loss": 3.8951,
+      "step": 1684480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2411755919135883e-05,
+      "loss": 3.8982,
+      "step": 1684992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2403386350429092e-05,
+      "loss": 3.8879,
+      "step": 1685504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2395000402918572e-05,
+      "loss": 3.8864,
+      "step": 1686016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2386614455408052e-05,
+      "loss": 3.8961,
+      "step": 1686528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2378228507897532e-05,
+      "loss": 3.8837,
+      "step": 1687040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2369842560387015e-05,
+      "loss": 3.8907,
+      "step": 1687552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2361456612876492e-05,
+      "loss": 3.8954,
+      "step": 1688064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2353070665365972e-05,
+      "loss": 3.8881,
+      "step": 1688576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.234468471785545e-05,
+      "loss": 3.9129,
+      "step": 1689088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.233629877034493e-05,
+      "loss": 3.9016,
+      "step": 1689600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2327929201638144e-05,
+      "loss": 3.8935,
+      "step": 1690112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2319543254127624e-05,
+      "loss": 3.8978,
+      "step": 1690624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2311157306617104e-05,
+      "loss": 3.9088,
+      "step": 1691136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2302771359106584e-05,
+      "loss": 3.8877,
+      "step": 1691648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2294401790399793e-05,
+      "loss": 3.9003,
+      "step": 1692160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2286015842889273e-05,
+      "loss": 3.8901,
+      "step": 1692672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2277629895378753e-05,
+      "loss": 3.8915,
+      "step": 1693184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2269243947868236e-05,
+      "loss": 3.8878,
+      "step": 1693696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2260858000357716e-05,
+      "loss": 3.896,
+      "step": 1694208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2252488431650925e-05,
+      "loss": 3.9047,
+      "step": 1694720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2244102484140405e-05,
+      "loss": 3.8989,
+      "step": 1695232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2235732915433614e-05,
+      "loss": 3.9037,
+      "step": 1695744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2227346967923098e-05,
+      "loss": 3.9013,
+      "step": 1696256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.221897739921631e-05,
+      "loss": 3.8915,
+      "step": 1696768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2210591451705787e-05,
+      "loss": 3.8983,
+      "step": 1697280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2202205504195267e-05,
+      "loss": 3.8887,
+      "step": 1697792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2193819556684747e-05,
+      "loss": 3.8886,
+      "step": 1698304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2185433609174227e-05,
+      "loss": 3.887,
+      "step": 1698816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2177047661663707e-05,
+      "loss": 3.886,
+      "step": 1699328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2168661714153187e-05,
+      "loss": 3.8952,
+      "step": 1699840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.216027576664267e-05,
+      "loss": 3.8992,
+      "step": 1700352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2151889819132147e-05,
+      "loss": 3.8898,
+      "step": 1700864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2143503871621626e-05,
+      "loss": 3.89,
+      "step": 1701376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2135117924111106e-05,
+      "loss": 3.901,
+      "step": 1701888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2126731976600586e-05,
+      "loss": 3.8907,
+      "step": 1702400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.21183624078938e-05,
+      "loss": 3.8986,
+      "step": 1702912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.210997646038328e-05,
+      "loss": 3.8913,
+      "step": 1703424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.210159051287276e-05,
+      "loss": 3.8754,
+      "step": 1703936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.209320456536224e-05,
+      "loss": 3.9058,
+      "step": 1704448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2084834996655448e-05,
+      "loss": 3.8876,
+      "step": 1704960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2076449049144928e-05,
+      "loss": 3.898,
+      "step": 1705472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2068063101634408e-05,
+      "loss": 3.885,
+      "step": 1705984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.205969353292762e-05,
+      "loss": 3.8893,
+      "step": 1706496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.20513075854171e-05,
+      "loss": 3.8847,
+      "step": 1707008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.204292163790658e-05,
+      "loss": 3.8969,
+      "step": 1707520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.203453569039606e-05,
+      "loss": 3.8833,
+      "step": 1708032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.202614974288554e-05,
+      "loss": 3.8898,
+      "step": 1708544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2017780174178753e-05,
+      "loss": 3.8988,
+      "step": 1709056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2009394226668232e-05,
+      "loss": 3.8816,
+      "step": 1709568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2001008279157712e-05,
+      "loss": 3.8834,
+      "step": 1710080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1992622331647192e-05,
+      "loss": 3.9022,
+      "step": 1710592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1984236384136672e-05,
+      "loss": 3.8717,
+      "step": 1711104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.197585043662615e-05,
+      "loss": 3.8748,
+      "step": 1711616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.196746448911563e-05,
+      "loss": 3.8922,
+      "step": 1712128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.195907854160511e-05,
+      "loss": 3.8955,
+      "step": 1712640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1950708972898325e-05,
+      "loss": 3.8842,
+      "step": 1713152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.19423230253878e-05,
+      "loss": 3.8932,
+      "step": 1713664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.193393707787728e-05,
+      "loss": 3.8743,
+      "step": 1714176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.192555113036676e-05,
+      "loss": 3.883,
+      "step": 1714688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.191718156165997e-05,
+      "loss": 3.8956,
+      "step": 1715200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1908795614149454e-05,
+      "loss": 3.8924,
+      "step": 1715712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1900409666638934e-05,
+      "loss": 3.8967,
+      "step": 1716224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1892023719128414e-05,
+      "loss": 3.8995,
+      "step": 1716736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1883654150421623e-05,
+      "loss": 3.905,
+      "step": 1717248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1875284581714835e-05,
+      "loss": 3.8804,
+      "step": 1717760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1866898634204315e-05,
+      "loss": 3.8912,
+      "step": 1718272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1858512686693795e-05,
+      "loss": 3.8942,
+      "step": 1718784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1850126739183275e-05,
+      "loss": 3.8756,
+      "step": 1719296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1841740791672755e-05,
+      "loss": 3.8953,
+      "step": 1719808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1833354844162235e-05,
+      "loss": 3.8923,
+      "step": 1720320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1824968896651715e-05,
+      "loss": 3.8964,
+      "step": 1720832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1816582949141195e-05,
+      "loss": 3.8947,
+      "step": 1721344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1808213380434407e-05,
+      "loss": 3.8835,
+      "step": 1721856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1799827432923887e-05,
+      "loss": 3.887,
+      "step": 1722368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1791441485413367e-05,
+      "loss": 3.8921,
+      "step": 1722880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1783071916706576e-05,
+      "loss": 3.8949,
+      "step": 1723392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1774685969196056e-05,
+      "loss": 3.8917,
+      "step": 1723904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1766300021685536e-05,
+      "loss": 3.8994,
+      "step": 1724416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1757914074175016e-05,
+      "loss": 3.8988,
+      "step": 1724928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.17495281266645e-05,
+      "loss": 3.886,
+      "step": 1725440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.174114217915398e-05,
+      "loss": 3.8916,
+      "step": 1725952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1732756231643456e-05,
+      "loss": 3.8806,
+      "step": 1726464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1724370284132936e-05,
+      "loss": 3.902,
+      "step": 1726976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1715984336622416e-05,
+      "loss": 3.9021,
+      "step": 1727488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1707614767915625e-05,
+      "loss": 3.8874,
+      "step": 1728000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.169922882040511e-05,
+      "loss": 3.8854,
+      "step": 1728512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.169084287289459e-05,
+      "loss": 3.8815,
+      "step": 1729024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1682456925384068e-05,
+      "loss": 3.8976,
+      "step": 1729536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1674087356677277e-05,
+      "loss": 3.9019,
+      "step": 1730048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1665701409166757e-05,
+      "loss": 3.8862,
+      "step": 1730560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1657315461656237e-05,
+      "loss": 3.888,
+      "step": 1731072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1648929514145717e-05,
+      "loss": 3.8833,
+      "step": 1731584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.164055994543893e-05,
+      "loss": 3.8969,
+      "step": 1732096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.163217399792841e-05,
+      "loss": 3.8766,
+      "step": 1732608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.162378805041789e-05,
+      "loss": 3.8892,
+      "step": 1733120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.16154184817111e-05,
+      "loss": 3.8827,
+      "step": 1733632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.160703253420058e-05,
+      "loss": 3.8956,
+      "step": 1734144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1598646586690062e-05,
+      "loss": 3.8871,
+      "step": 1734656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1590260639179542e-05,
+      "loss": 3.8911,
+      "step": 1735168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.158189107047275e-05,
+      "loss": 3.8903,
+      "step": 1735680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.157350512296223e-05,
+      "loss": 3.8836,
+      "step": 1736192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.156511917545171e-05,
+      "loss": 3.8935,
+      "step": 1736704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.155673322794119e-05,
+      "loss": 3.891,
+      "step": 1737216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.154834728043067e-05,
+      "loss": 3.8972,
+      "step": 1737728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1539977711723883e-05,
+      "loss": 3.9005,
+      "step": 1738240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1531608143017096e-05,
+      "loss": 3.8869,
+      "step": 1738752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1523222195506572e-05,
+      "loss": 3.8905,
+      "step": 1739264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1514836247996052e-05,
+      "loss": 3.8825,
+      "step": 1739776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1506450300485532e-05,
+      "loss": 3.893,
+      "step": 1740288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1498064352975016e-05,
+      "loss": 3.8877,
+      "step": 1740800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1489678405464496e-05,
+      "loss": 3.8894,
+      "step": 1741312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1481292457953976e-05,
+      "loss": 3.8763,
+      "step": 1741824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1472906510443456e-05,
+      "loss": 3.8942,
+      "step": 1742336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1464536941736665e-05,
+      "loss": 3.8852,
+      "step": 1742848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1456150994226145e-05,
+      "loss": 3.8902,
+      "step": 1743360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1447765046715625e-05,
+      "loss": 3.8761,
+      "step": 1743872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1439379099205105e-05,
+      "loss": 3.8904,
+      "step": 1744384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1430993151694585e-05,
+      "loss": 3.8871,
+      "step": 1744896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1422607204184064e-05,
+      "loss": 3.8998,
+      "step": 1745408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1414237635477277e-05,
+      "loss": 3.8915,
+      "step": 1745920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1405851687966754e-05,
+      "loss": 3.8969,
+      "step": 1746432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1397465740456233e-05,
+      "loss": 3.8812,
+      "step": 1746944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1389079792945717e-05,
+      "loss": 3.8951,
+      "step": 1747456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1380693845435197e-05,
+      "loss": 3.8995,
+      "step": 1747968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1372324276728406e-05,
+      "loss": 3.8966,
+      "step": 1748480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1363938329217886e-05,
+      "loss": 3.895,
+      "step": 1748992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.13555687605111e-05,
+      "loss": 3.8994,
+      "step": 1749504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1347182813000578e-05,
+      "loss": 3.8889,
+      "step": 1750016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1338796865490058e-05,
+      "loss": 3.8974,
+      "step": 1750528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1330410917979538e-05,
+      "loss": 3.8821,
+      "step": 1751040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1322024970469018e-05,
+      "loss": 3.8917,
+      "step": 1751552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1313639022958498e-05,
+      "loss": 3.8843,
+      "step": 1752064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1305253075447978e-05,
+      "loss": 3.8907,
+      "step": 1752576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1296867127937458e-05,
+      "loss": 3.8846,
+      "step": 1753088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1288481180426938e-05,
+      "loss": 3.8826,
+      "step": 1753600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1280095232916418e-05,
+      "loss": 3.894,
+      "step": 1754112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.127172566420963e-05,
+      "loss": 3.881,
+      "step": 1754624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.126333971669911e-05,
+      "loss": 3.903,
+      "step": 1755136
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9876136779785156,
+      "eval_runtime": 297.9153,
+      "eval_samples_per_second": 1280.871,
+      "eval_steps_per_second": 40.028,
+      "step": 1755360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1254953769188587e-05,
+      "loss": 3.886,
+      "step": 1755648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1246567821678067e-05,
+      "loss": 3.8796,
+      "step": 1756160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1238181874167547e-05,
+      "loss": 3.891,
+      "step": 1756672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1229795926657027e-05,
+      "loss": 3.8909,
+      "step": 1757184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.122140997914651e-05,
+      "loss": 3.8897,
+      "step": 1757696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.121302403163599e-05,
+      "loss": 3.8868,
+      "step": 1758208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1204638084125467e-05,
+      "loss": 3.8809,
+      "step": 1758720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1196252136614947e-05,
+      "loss": 3.8835,
+      "step": 1759232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1187866189104427e-05,
+      "loss": 3.8915,
+      "step": 1759744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1179480241593907e-05,
+      "loss": 3.8996,
+      "step": 1760256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1171094294083386e-05,
+      "loss": 3.8885,
+      "step": 1760768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.116270834657287e-05,
+      "loss": 3.8893,
+      "step": 1761280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.115432239906235e-05,
+      "loss": 3.8858,
+      "step": 1761792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.114595283035556e-05,
+      "loss": 3.879,
+      "step": 1762304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.113756688284504e-05,
+      "loss": 3.8874,
+      "step": 1762816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.112918093533452e-05,
+      "loss": 3.8752,
+      "step": 1763328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1120794987824e-05,
+      "loss": 3.8846,
+      "step": 1763840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.111240904031348e-05,
+      "loss": 3.8901,
+      "step": 1764352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.110403947160669e-05,
+      "loss": 3.8823,
+      "step": 1764864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.109565352409617e-05,
+      "loss": 3.9042,
+      "step": 1765376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1087267576585648e-05,
+      "loss": 3.8982,
+      "step": 1765888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1078881629075128e-05,
+      "loss": 3.8865,
+      "step": 1766400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1070495681564608e-05,
+      "loss": 3.895,
+      "step": 1766912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1062109734054088e-05,
+      "loss": 3.9034,
+      "step": 1767424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.105372378654357e-05,
+      "loss": 3.8801,
+      "step": 1767936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.104533783903305e-05,
+      "loss": 3.894,
+      "step": 1768448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.103695189152253e-05,
+      "loss": 3.8852,
+      "step": 1768960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1028565944012007e-05,
+      "loss": 3.885,
+      "step": 1769472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1020179996501487e-05,
+      "loss": 3.8803,
+      "step": 1769984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1011794048990967e-05,
+      "loss": 3.8918,
+      "step": 1770496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1003408101480447e-05,
+      "loss": 3.8975,
+      "step": 1771008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.099503853277366e-05,
+      "loss": 3.8944,
+      "step": 1771520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.098665258526314e-05,
+      "loss": 3.8977,
+      "step": 1772032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.097826663775262e-05,
+      "loss": 3.8881,
+      "step": 1772544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.096989706904583e-05,
+      "loss": 3.8857,
+      "step": 1773056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.096151112153531e-05,
+      "loss": 3.8935,
+      "step": 1773568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.095312517402479e-05,
+      "loss": 3.8844,
+      "step": 1774080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0944739226514272e-05,
+      "loss": 3.8778,
+      "step": 1774592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0936353279003752e-05,
+      "loss": 3.8817,
+      "step": 1775104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0927967331493232e-05,
+      "loss": 3.8789,
+      "step": 1775616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0919581383982712e-05,
+      "loss": 3.8876,
+      "step": 1776128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.091121181527592e-05,
+      "loss": 3.8941,
+      "step": 1776640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.09028258677654e-05,
+      "loss": 3.8864,
+      "step": 1777152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.089443992025488e-05,
+      "loss": 3.8831,
+      "step": 1777664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.088605397274436e-05,
+      "loss": 3.8964,
+      "step": 1778176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.087766802523384e-05,
+      "loss": 3.8886,
+      "step": 1778688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.086928207772332e-05,
+      "loss": 3.89,
+      "step": 1779200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.08608961302128e-05,
+      "loss": 3.8865,
+      "step": 1779712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.085251018270228e-05,
+      "loss": 3.8639,
+      "step": 1780224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0844140613995493e-05,
+      "loss": 3.9031,
+      "step": 1780736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0835771045288706e-05,
+      "loss": 3.8817,
+      "step": 1781248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0827385097778186e-05,
+      "loss": 3.8945,
+      "step": 1781760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0818999150267662e-05,
+      "loss": 3.8793,
+      "step": 1782272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0810613202757142e-05,
+      "loss": 3.8854,
+      "step": 1782784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0802227255246622e-05,
+      "loss": 3.8781,
+      "step": 1783296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0793841307736102e-05,
+      "loss": 3.8885,
+      "step": 1783808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0785455360225582e-05,
+      "loss": 3.8779,
+      "step": 1784320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0777069412715065e-05,
+      "loss": 3.8833,
+      "step": 1784832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0768699844008274e-05,
+      "loss": 3.8942,
+      "step": 1785344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0760313896497754e-05,
+      "loss": 3.8779,
+      "step": 1785856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0751927948987234e-05,
+      "loss": 3.8738,
+      "step": 1786368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0743542001476714e-05,
+      "loss": 3.8934,
+      "step": 1786880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0735172432769927e-05,
+      "loss": 3.8693,
+      "step": 1787392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0726802864063136e-05,
+      "loss": 3.8681,
+      "step": 1787904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0718416916552616e-05,
+      "loss": 3.8857,
+      "step": 1788416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0710030969042096e-05,
+      "loss": 3.8869,
+      "step": 1788928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0701645021531576e-05,
+      "loss": 3.8787,
+      "step": 1789440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0693259074021056e-05,
+      "loss": 3.8881,
+      "step": 1789952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0684873126510536e-05,
+      "loss": 3.8671,
+      "step": 1790464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0676503557803748e-05,
+      "loss": 3.8797,
+      "step": 1790976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0668117610293228e-05,
+      "loss": 3.8889,
+      "step": 1791488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0659731662782708e-05,
+      "loss": 3.8843,
+      "step": 1792000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0651345715272188e-05,
+      "loss": 3.8941,
+      "step": 1792512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0642959767761665e-05,
+      "loss": 3.8914,
+      "step": 1793024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0634573820251148e-05,
+      "loss": 3.8962,
+      "step": 1793536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.062620425154436e-05,
+      "loss": 3.8729,
+      "step": 1794048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.061781830403384e-05,
+      "loss": 3.8887,
+      "step": 1794560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0609432356523317e-05,
+      "loss": 3.8866,
+      "step": 1795072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0601046409012797e-05,
+      "loss": 3.8707,
+      "step": 1795584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0592660461502277e-05,
+      "loss": 3.8882,
+      "step": 1796096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0584274513991757e-05,
+      "loss": 3.889,
+      "step": 1796608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.057590494528497e-05,
+      "loss": 3.8911,
+      "step": 1797120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.056751899777445e-05,
+      "loss": 3.8944,
+      "step": 1797632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.055913305026393e-05,
+      "loss": 3.8747,
+      "step": 1798144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0550763481557138e-05,
+      "loss": 3.8804,
+      "step": 1798656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0542377534046618e-05,
+      "loss": 3.8864,
+      "step": 1799168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.05339915865361e-05,
+      "loss": 3.8873,
+      "step": 1799680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.052560563902558e-05,
+      "loss": 3.8868,
+      "step": 1800192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.051721969151506e-05,
+      "loss": 3.8893,
+      "step": 1800704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.050883374400454e-05,
+      "loss": 3.8956,
+      "step": 1801216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.050044779649402e-05,
+      "loss": 3.8835,
+      "step": 1801728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0492061848983498e-05,
+      "loss": 3.8843,
+      "step": 1802240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0483675901472978e-05,
+      "loss": 3.8767,
+      "step": 1802752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.047530633276619e-05,
+      "loss": 3.8953,
+      "step": 1803264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.046692038525567e-05,
+      "loss": 3.8967,
+      "step": 1803776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.045853443774515e-05,
+      "loss": 3.8835,
+      "step": 1804288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.045014849023463e-05,
+      "loss": 3.8784,
+      "step": 1804800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.044176254272411e-05,
+      "loss": 3.8777,
+      "step": 1805312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.043339297401732e-05,
+      "loss": 3.8889,
+      "step": 1805824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0425007026506803e-05,
+      "loss": 3.8998,
+      "step": 1806336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0416621078996283e-05,
+      "loss": 3.8804,
+      "step": 1806848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0408235131485763e-05,
+      "loss": 3.8834,
+      "step": 1807360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0399849183975243e-05,
+      "loss": 3.8796,
+      "step": 1807872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0391463236464722e-05,
+      "loss": 3.8853,
+      "step": 1808384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.03830772889542e-05,
+      "loss": 3.8738,
+      "step": 1808896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.037469134144368e-05,
+      "loss": 3.8864,
+      "step": 1809408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0366321772736895e-05,
+      "loss": 3.8706,
+      "step": 1809920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0357935825226375e-05,
+      "loss": 3.8883,
+      "step": 1810432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.034954987771585e-05,
+      "loss": 3.884,
+      "step": 1810944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.034116393020533e-05,
+      "loss": 3.8825,
+      "step": 1811456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0332794361498544e-05,
+      "loss": 3.8837,
+      "step": 1811968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0324424792791756e-05,
+      "loss": 3.8851,
+      "step": 1812480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0316038845281236e-05,
+      "loss": 3.8841,
+      "step": 1812992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0307652897770716e-05,
+      "loss": 3.8868,
+      "step": 1813504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0299266950260196e-05,
+      "loss": 3.8897,
+      "step": 1814016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0290881002749673e-05,
+      "loss": 3.8973,
+      "step": 1814528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0282495055239153e-05,
+      "loss": 3.8829,
+      "step": 1815040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0274125486532365e-05,
+      "loss": 3.8847,
+      "step": 1815552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.026573953902185e-05,
+      "loss": 3.8771,
+      "step": 1816064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0257353591511325e-05,
+      "loss": 3.8845,
+      "step": 1816576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0248967644000805e-05,
+      "loss": 3.8839,
+      "step": 1817088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0240581696490285e-05,
+      "loss": 3.8833,
+      "step": 1817600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0232195748979765e-05,
+      "loss": 3.8724,
+      "step": 1818112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0223809801469245e-05,
+      "loss": 3.8846,
+      "step": 1818624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0215423853958725e-05,
+      "loss": 3.8841,
+      "step": 1819136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0207037906448205e-05,
+      "loss": 3.8834,
+      "step": 1819648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0198668337741417e-05,
+      "loss": 3.868,
+      "step": 1820160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0190282390230897e-05,
+      "loss": 3.8882,
+      "step": 1820672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0181896442720377e-05,
+      "loss": 3.8793,
+      "step": 1821184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0173510495209854e-05,
+      "loss": 3.8935,
+      "step": 1821696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0165124547699334e-05,
+      "loss": 3.8868,
+      "step": 1822208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0156738600188814e-05,
+      "loss": 3.8902,
+      "step": 1822720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0148352652678297e-05,
+      "loss": 3.8763,
+      "step": 1823232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0139983083971506e-05,
+      "loss": 3.8911,
+      "step": 1823744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0131597136460986e-05,
+      "loss": 3.891,
+      "step": 1824256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0123211188950466e-05,
+      "loss": 3.8905,
+      "step": 1824768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0114825241439946e-05,
+      "loss": 3.8923,
+      "step": 1825280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0106439293929426e-05,
+      "loss": 3.891,
+      "step": 1825792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.009806972522264e-05,
+      "loss": 3.8845,
+      "step": 1826304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.008968377771212e-05,
+      "loss": 3.8917,
+      "step": 1826816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.00812978302016e-05,
+      "loss": 3.8766,
+      "step": 1827328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.007291188269108e-05,
+      "loss": 3.8881,
+      "step": 1827840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0064525935180558e-05,
+      "loss": 3.8797,
+      "step": 1828352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0056139987670035e-05,
+      "loss": 3.8881,
+      "step": 1828864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.004777041896325e-05,
+      "loss": 3.8762,
+      "step": 1829376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.003938447145273e-05,
+      "loss": 3.8795,
+      "step": 1829888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.003099852394221e-05,
+      "loss": 3.885,
+      "step": 1830400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0022612576431687e-05,
+      "loss": 3.8793,
+      "step": 1830912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0014226628921167e-05,
+      "loss": 3.8964,
+      "step": 1831424
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9853124618530273,
+      "eval_runtime": 295.7922,
+      "eval_samples_per_second": 1290.064,
+      "eval_steps_per_second": 40.315,
+      "step": 1831680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0005840681410647e-05,
+      "loss": 3.8745,
+      "step": 1831936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.999747111270386e-05,
+      "loss": 3.8782,
+      "step": 1832448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.998908516519334e-05,
+      "loss": 3.8819,
+      "step": 1832960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.998069921768282e-05,
+      "loss": 3.884,
+      "step": 1833472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.99723132701723e-05,
+      "loss": 3.8864,
+      "step": 1833984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.996392732266178e-05,
+      "loss": 3.8789,
+      "step": 1834496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.995554137515126e-05,
+      "loss": 3.8839,
+      "step": 1835008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.994715542764074e-05,
+      "loss": 3.8776,
+      "step": 1835520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.993876948013022e-05,
+      "loss": 3.8817,
+      "step": 1836032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9930399911423432e-05,
+      "loss": 3.895,
+      "step": 1836544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9922013963912912e-05,
+      "loss": 3.8813,
+      "step": 1837056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.991364439520612e-05,
+      "loss": 3.8829,
+      "step": 1837568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.99052584476956e-05,
+      "loss": 3.8796,
+      "step": 1838080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.989687250018508e-05,
+      "loss": 3.8725,
+      "step": 1838592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.988848655267456e-05,
+      "loss": 3.8842,
+      "step": 1839104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9880100605164044e-05,
+      "loss": 3.8698,
+      "step": 1839616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.987171465765352e-05,
+      "loss": 3.8819,
+      "step": 1840128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9863328710143e-05,
+      "loss": 3.8873,
+      "step": 1840640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.985494276263248e-05,
+      "loss": 3.8727,
+      "step": 1841152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.984657319392569e-05,
+      "loss": 3.8974,
+      "step": 1841664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9838187246415173e-05,
+      "loss": 3.8974,
+      "step": 1842176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9829801298904653e-05,
+      "loss": 3.8824,
+      "step": 1842688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9821415351394133e-05,
+      "loss": 3.8817,
+      "step": 1843200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9813045782687342e-05,
+      "loss": 3.8936,
+      "step": 1843712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9804659835176822e-05,
+      "loss": 3.8821,
+      "step": 1844224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9796273887666302e-05,
+      "loss": 3.8843,
+      "step": 1844736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9787887940155782e-05,
+      "loss": 3.8797,
+      "step": 1845248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9779501992645265e-05,
+      "loss": 3.8796,
+      "step": 1845760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9771132423938474e-05,
+      "loss": 3.8804,
+      "step": 1846272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9762746476427954e-05,
+      "loss": 3.8825,
+      "step": 1846784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9754360528917434e-05,
+      "loss": 3.8882,
+      "step": 1847296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9745990960210643e-05,
+      "loss": 3.8924,
+      "step": 1847808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9737605012700127e-05,
+      "loss": 3.8909,
+      "step": 1848320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9729219065189607e-05,
+      "loss": 3.8846,
+      "step": 1848832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9720833117679087e-05,
+      "loss": 3.8791,
+      "step": 1849344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9712463548972296e-05,
+      "loss": 3.8886,
+      "step": 1849856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9704077601461776e-05,
+      "loss": 3.8817,
+      "step": 1850368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9695691653951256e-05,
+      "loss": 3.8732,
+      "step": 1850880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9687305706440735e-05,
+      "loss": 3.8762,
+      "step": 1851392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9678919758930215e-05,
+      "loss": 3.8713,
+      "step": 1851904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9670533811419695e-05,
+      "loss": 3.8812,
+      "step": 1852416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9662147863909175e-05,
+      "loss": 3.8856,
+      "step": 1852928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9653761916398655e-05,
+      "loss": 3.8848,
+      "step": 1853440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9645375968888135e-05,
+      "loss": 3.8801,
+      "step": 1853952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9636990021377615e-05,
+      "loss": 3.8844,
+      "step": 1854464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9628604073867095e-05,
+      "loss": 3.8874,
+      "step": 1854976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9620218126356575e-05,
+      "loss": 3.8818,
+      "step": 1855488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9611848557649788e-05,
+      "loss": 3.8854,
+      "step": 1856000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9603462610139268e-05,
+      "loss": 3.8576,
+      "step": 1856512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9595076662628748e-05,
+      "loss": 3.8959,
+      "step": 1857024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9586690715118227e-05,
+      "loss": 3.8757,
+      "step": 1857536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.957833752521517e-05,
+      "loss": 3.8921,
+      "step": 1858048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.956995157770465e-05,
+      "loss": 3.8729,
+      "step": 1858560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.956156563019413e-05,
+      "loss": 3.877,
+      "step": 1859072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.955317968268361e-05,
+      "loss": 3.8723,
+      "step": 1859584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.954479373517309e-05,
+      "loss": 3.8812,
+      "step": 1860096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.953640778766257e-05,
+      "loss": 3.8748,
+      "step": 1860608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.952802184015205e-05,
+      "loss": 3.876,
+      "step": 1861120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951963589264153e-05,
+      "loss": 3.8869,
+      "step": 1861632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951124994513101e-05,
+      "loss": 3.8755,
+      "step": 1862144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.950288037642422e-05,
+      "loss": 3.8664,
+      "step": 1862656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.94944944289137e-05,
+      "loss": 3.8923,
+      "step": 1863168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.948612486020691e-05,
+      "loss": 3.864,
+      "step": 1863680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.947773891269639e-05,
+      "loss": 3.8609,
+      "step": 1864192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9469352965185874e-05,
+      "loss": 3.8812,
+      "step": 1864704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.946096701767535e-05,
+      "loss": 3.8819,
+      "step": 1865216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.945258107016483e-05,
+      "loss": 3.8743,
+      "step": 1865728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.944419512265431e-05,
+      "loss": 3.8795,
+      "step": 1866240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9435825553947523e-05,
+      "loss": 3.8616,
+      "step": 1866752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9427439606437e-05,
+      "loss": 3.8753,
+      "step": 1867264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9419053658926482e-05,
+      "loss": 3.8855,
+      "step": 1867776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9410667711415962e-05,
+      "loss": 3.877,
+      "step": 1868288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9402298142709175e-05,
+      "loss": 3.8885,
+      "step": 1868800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.939391219519865e-05,
+      "loss": 3.8823,
+      "step": 1869312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.938552624768813e-05,
+      "loss": 3.8909,
+      "step": 1869824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9377156678981344e-05,
+      "loss": 3.868,
+      "step": 1870336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9368770731470824e-05,
+      "loss": 3.8825,
+      "step": 1870848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9360384783960304e-05,
+      "loss": 3.8775,
+      "step": 1871360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9351998836449784e-05,
+      "loss": 3.8704,
+      "step": 1871872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9343612888939264e-05,
+      "loss": 3.8807,
+      "step": 1872384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9335243320232473e-05,
+      "loss": 3.8852,
+      "step": 1872896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9326857372721953e-05,
+      "loss": 3.8822,
+      "step": 1873408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9318471425211436e-05,
+      "loss": 3.8908,
+      "step": 1873920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9310085477700916e-05,
+      "loss": 3.8666,
+      "step": 1874432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9301699530190396e-05,
+      "loss": 3.8726,
+      "step": 1874944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9293313582679876e-05,
+      "loss": 3.8862,
+      "step": 1875456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9284927635169356e-05,
+      "loss": 3.8819,
+      "step": 1875968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9276541687658833e-05,
+      "loss": 3.8819,
+      "step": 1876480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9268155740148312e-05,
+      "loss": 3.8886,
+      "step": 1876992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.925978617144153e-05,
+      "loss": 3.8834,
+      "step": 1877504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9251400223931005e-05,
+      "loss": 3.8813,
+      "step": 1878016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9243014276420485e-05,
+      "loss": 3.879,
+      "step": 1878528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9234628328909965e-05,
+      "loss": 3.8731,
+      "step": 1879040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9226258760203177e-05,
+      "loss": 3.8908,
+      "step": 1879552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9217872812692654e-05,
+      "loss": 3.8838,
+      "step": 1880064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9209486865182137e-05,
+      "loss": 3.882,
+      "step": 1880576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9201100917671617e-05,
+      "loss": 3.876,
+      "step": 1881088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9192714970161097e-05,
+      "loss": 3.8698,
+      "step": 1881600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9184329022650577e-05,
+      "loss": 3.8814,
+      "step": 1882112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9175959453943786e-05,
+      "loss": 3.892,
+      "step": 1882624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9167573506433266e-05,
+      "loss": 3.8769,
+      "step": 1883136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9159187558922746e-05,
+      "loss": 3.8796,
+      "step": 1883648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.915080161141223e-05,
+      "loss": 3.8723,
+      "step": 1884160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.914241566390171e-05,
+      "loss": 3.8757,
+      "step": 1884672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9134029716391186e-05,
+      "loss": 3.8774,
+      "step": 1885184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9125643768880666e-05,
+      "loss": 3.8744,
+      "step": 1885696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9117257821370146e-05,
+      "loss": 3.8678,
+      "step": 1886208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.910890463146709e-05,
+      "loss": 3.8846,
+      "step": 1886720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.910051868395657e-05,
+      "loss": 3.8803,
+      "step": 1887232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.909213273644605e-05,
+      "loss": 3.8765,
+      "step": 1887744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.908376316773926e-05,
+      "loss": 3.8722,
+      "step": 1888256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.907537722022874e-05,
+      "loss": 3.8833,
+      "step": 1888768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9067024030325685e-05,
+      "loss": 3.8804,
+      "step": 1889280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9058638082815165e-05,
+      "loss": 3.8818,
+      "step": 1889792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9050252135304645e-05,
+      "loss": 3.885,
+      "step": 1890304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9041866187794125e-05,
+      "loss": 3.8918,
+      "step": 1890816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.90334802402836e-05,
+      "loss": 3.8756,
+      "step": 1891328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.902509429277308e-05,
+      "loss": 3.8838,
+      "step": 1891840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.901670834526256e-05,
+      "loss": 3.8692,
+      "step": 1892352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9008338776555774e-05,
+      "loss": 3.8817,
+      "step": 1892864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8999952829045254e-05,
+      "loss": 3.8789,
+      "step": 1893376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8991566881534734e-05,
+      "loss": 3.8743,
+      "step": 1893888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8983180934024214e-05,
+      "loss": 3.87,
+      "step": 1894400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8974794986513694e-05,
+      "loss": 3.8794,
+      "step": 1894912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8966409039003173e-05,
+      "loss": 3.8801,
+      "step": 1895424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8958023091492653e-05,
+      "loss": 3.8789,
+      "step": 1895936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8949637143982133e-05,
+      "loss": 3.8643,
+      "step": 1896448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8941251196471613e-05,
+      "loss": 3.8786,
+      "step": 1896960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8932865248961093e-05,
+      "loss": 3.8743,
+      "step": 1897472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8924479301450573e-05,
+      "loss": 3.8884,
+      "step": 1897984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8916093353940053e-05,
+      "loss": 3.8833,
+      "step": 1898496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8907707406429533e-05,
+      "loss": 3.8803,
+      "step": 1899008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8899321458919013e-05,
+      "loss": 3.8725,
+      "step": 1899520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8890935511408493e-05,
+      "loss": 3.887,
+      "step": 1900032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8882549563897973e-05,
+      "loss": 3.8862,
+      "step": 1900544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8874163616387453e-05,
+      "loss": 3.886,
+      "step": 1901056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8865777668876933e-05,
+      "loss": 3.8846,
+      "step": 1901568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8857391721366413e-05,
+      "loss": 3.8849,
+      "step": 1902080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8849022152659622e-05,
+      "loss": 3.88,
+      "step": 1902592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8840636205149105e-05,
+      "loss": 3.886,
+      "step": 1903104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8832250257638585e-05,
+      "loss": 3.8717,
+      "step": 1903616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8823864310128065e-05,
+      "loss": 3.8823,
+      "step": 1904128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8815478362617545e-05,
+      "loss": 3.8782,
+      "step": 1904640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8807092415107022e-05,
+      "loss": 3.8802,
+      "step": 1905152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8798706467596502e-05,
+      "loss": 3.871,
+      "step": 1905664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8790336898889714e-05,
+      "loss": 3.8738,
+      "step": 1906176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8781967330182927e-05,
+      "loss": 3.8785,
+      "step": 1906688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8773581382672407e-05,
+      "loss": 3.8771,
+      "step": 1907200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8765211813965616e-05,
+      "loss": 3.8904,
+      "step": 1907712
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9836716651916504,
+      "eval_runtime": 295.6621,
+      "eval_samples_per_second": 1290.632,
+      "eval_steps_per_second": 40.333,
+      "step": 1908000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8756825866455096e-05,
+      "loss": 3.8677,
+      "step": 1908224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8748439918944576e-05,
+      "loss": 3.8733,
+      "step": 1908736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8740053971434056e-05,
+      "loss": 3.874,
+      "step": 1909248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.873166802392354e-05,
+      "loss": 3.886,
+      "step": 1909760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.872328207641302e-05,
+      "loss": 3.8821,
+      "step": 1910272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8714896128902495e-05,
+      "loss": 3.8714,
+      "step": 1910784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8706510181391975e-05,
+      "loss": 3.879,
+      "step": 1911296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8698124233881455e-05,
+      "loss": 3.8707,
+      "step": 1911808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8689738286370935e-05,
+      "loss": 3.8733,
+      "step": 1912320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8681352338860415e-05,
+      "loss": 3.8934,
+      "step": 1912832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.86729663913499e-05,
+      "loss": 3.873,
+      "step": 1913344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.866458044383938e-05,
+      "loss": 3.8737,
+      "step": 1913856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8656194496328855e-05,
+      "loss": 3.8849,
+      "step": 1914368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8647808548818335e-05,
+      "loss": 3.8601,
+      "step": 1914880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8639422601307815e-05,
+      "loss": 3.8762,
+      "step": 1915392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8631036653797295e-05,
+      "loss": 3.8651,
+      "step": 1915904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8622667085090508e-05,
+      "loss": 3.8785,
+      "step": 1916416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8614281137579987e-05,
+      "loss": 3.8778,
+      "step": 1916928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.86059115688732e-05,
+      "loss": 3.8695,
+      "step": 1917440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8597525621362677e-05,
+      "loss": 3.8901,
+      "step": 1917952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8589139673852156e-05,
+      "loss": 3.8896,
+      "step": 1918464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8580753726341636e-05,
+      "loss": 3.8808,
+      "step": 1918976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8572367778831116e-05,
+      "loss": 3.882,
+      "step": 1919488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.85639818313206e-05,
+      "loss": 3.8829,
+      "step": 1920000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.855559588381008e-05,
+      "loss": 3.8778,
+      "step": 1920512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8547209936299556e-05,
+      "loss": 3.8747,
+      "step": 1921024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8538823988789036e-05,
+      "loss": 3.8773,
+      "step": 1921536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8530438041278516e-05,
+      "loss": 3.8766,
+      "step": 1922048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8522052093767996e-05,
+      "loss": 3.8722,
+      "step": 1922560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.851368252506121e-05,
+      "loss": 3.8773,
+      "step": 1923072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.850529657755069e-05,
+      "loss": 3.8794,
+      "step": 1923584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.849691063004017e-05,
+      "loss": 3.8911,
+      "step": 1924096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.848852468252965e-05,
+      "loss": 3.8836,
+      "step": 1924608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.848013873501913e-05,
+      "loss": 3.8812,
+      "step": 1925120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.847175278750861e-05,
+      "loss": 3.8731,
+      "step": 1925632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8463383218801817e-05,
+      "loss": 3.8789,
+      "step": 1926144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.84549972712913e-05,
+      "loss": 3.8809,
+      "step": 1926656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.844661132378078e-05,
+      "loss": 3.871,
+      "step": 1927168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.843824175507399e-05,
+      "loss": 3.8715,
+      "step": 1927680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.842985580756347e-05,
+      "loss": 3.867,
+      "step": 1928192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.842146986005295e-05,
+      "loss": 3.8788,
+      "step": 1928704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.841308391254243e-05,
+      "loss": 3.8769,
+      "step": 1929216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.840469796503191e-05,
+      "loss": 3.8797,
+      "step": 1929728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.839631201752139e-05,
+      "loss": 3.8764,
+      "step": 1930240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.838792607001087e-05,
+      "loss": 3.8822,
+      "step": 1930752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.837954012250035e-05,
+      "loss": 3.8747,
+      "step": 1931264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.837115417498983e-05,
+      "loss": 3.8777,
+      "step": 1931776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.836276822747931e-05,
+      "loss": 3.8838,
+      "step": 1932288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8354398658772522e-05,
+      "loss": 3.8517,
+      "step": 1932800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8346012711262002e-05,
+      "loss": 3.8856,
+      "step": 1933312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8337626763751482e-05,
+      "loss": 3.8726,
+      "step": 1933824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8329240816240962e-05,
+      "loss": 3.8869,
+      "step": 1934336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8320854868730442e-05,
+      "loss": 3.8694,
+      "step": 1934848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.831246892121992e-05,
+      "loss": 3.8733,
+      "step": 1935360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.83040829737094e-05,
+      "loss": 3.8652,
+      "step": 1935872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8295697026198878e-05,
+      "loss": 3.8756,
+      "step": 1936384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.828731107868836e-05,
+      "loss": 3.8751,
+      "step": 1936896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.827894150998157e-05,
+      "loss": 3.8683,
+      "step": 1937408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.827055556247105e-05,
+      "loss": 3.8832,
+      "step": 1937920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8262185993764263e-05,
+      "loss": 3.8739,
+      "step": 1938432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8253800046253743e-05,
+      "loss": 3.8563,
+      "step": 1938944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8245414098743223e-05,
+      "loss": 3.8854,
+      "step": 1939456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8237028151232703e-05,
+      "loss": 3.8635,
+      "step": 1939968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8228642203722183e-05,
+      "loss": 3.8573,
+      "step": 1940480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8220272635015392e-05,
+      "loss": 3.8773,
+      "step": 1940992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8211886687504872e-05,
+      "loss": 3.875,
+      "step": 1941504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8203500739994352e-05,
+      "loss": 3.8659,
+      "step": 1942016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8195114792483832e-05,
+      "loss": 3.8826,
+      "step": 1942528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8186728844973315e-05,
+      "loss": 3.8584,
+      "step": 1943040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8178359276266524e-05,
+      "loss": 3.8655,
+      "step": 1943552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8169973328756004e-05,
+      "loss": 3.8784,
+      "step": 1944064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8161587381245484e-05,
+      "loss": 3.8724,
+      "step": 1944576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8153201433734964e-05,
+      "loss": 3.8879,
+      "step": 1945088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8144815486224444e-05,
+      "loss": 3.876,
+      "step": 1945600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8136429538713924e-05,
+      "loss": 3.8866,
+      "step": 1946112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8128059970007137e-05,
+      "loss": 3.8669,
+      "step": 1946624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8119674022496617e-05,
+      "loss": 3.8837,
+      "step": 1947136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8111288074986097e-05,
+      "loss": 3.8669,
+      "step": 1947648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8102902127475573e-05,
+      "loss": 3.8697,
+      "step": 1948160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8094516179965053e-05,
+      "loss": 3.874,
+      "step": 1948672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8086130232454533e-05,
+      "loss": 3.8776,
+      "step": 1949184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8077744284944016e-05,
+      "loss": 3.8805,
+      "step": 1949696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8069358337433496e-05,
+      "loss": 3.8878,
+      "step": 1950208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8060972389922976e-05,
+      "loss": 3.8608,
+      "step": 1950720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8052602821216185e-05,
+      "loss": 3.8667,
+      "step": 1951232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8044216873705665e-05,
+      "loss": 3.8796,
+      "step": 1951744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8035830926195145e-05,
+      "loss": 3.8792,
+      "step": 1952256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8027444978684625e-05,
+      "loss": 3.8742,
+      "step": 1952768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8019075409977838e-05,
+      "loss": 3.8856,
+      "step": 1953280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8010689462467318e-05,
+      "loss": 3.8769,
+      "step": 1953792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8002303514956798e-05,
+      "loss": 3.8799,
+      "step": 1954304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7993917567446278e-05,
+      "loss": 3.8761,
+      "step": 1954816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7985547998739487e-05,
+      "loss": 3.8686,
+      "step": 1955328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.797716205122897e-05,
+      "loss": 3.8836,
+      "step": 1955840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.796877610371845e-05,
+      "loss": 3.8826,
+      "step": 1956352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.796039015620793e-05,
+      "loss": 3.8756,
+      "step": 1956864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7952036966304868e-05,
+      "loss": 3.8736,
+      "step": 1957376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7943651018794348e-05,
+      "loss": 3.8629,
+      "step": 1957888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.793526507128383e-05,
+      "loss": 3.875,
+      "step": 1958400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7926895502577044e-05,
+      "loss": 3.8875,
+      "step": 1958912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.791850955506652e-05,
+      "loss": 3.8729,
+      "step": 1959424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7910123607556e-05,
+      "loss": 3.8755,
+      "step": 1959936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.790173766004548e-05,
+      "loss": 3.8692,
+      "step": 1960448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.789335171253496e-05,
+      "loss": 3.8682,
+      "step": 1960960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.788496576502444e-05,
+      "loss": 3.8733,
+      "step": 1961472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7876579817513924e-05,
+      "loss": 3.8664,
+      "step": 1961984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7868193870003404e-05,
+      "loss": 3.8645,
+      "step": 1962496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7859824301296613e-05,
+      "loss": 3.8792,
+      "step": 1963008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7851438353786093e-05,
+      "loss": 3.8788,
+      "step": 1963520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7843052406275573e-05,
+      "loss": 3.866,
+      "step": 1964032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.783466645876505e-05,
+      "loss": 3.8704,
+      "step": 1964544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7826280511254533e-05,
+      "loss": 3.874,
+      "step": 1965056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7817894563744013e-05,
+      "loss": 3.8788,
+      "step": 1965568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7809524995037225e-05,
+      "loss": 3.8733,
+      "step": 1966080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.78011390475267e-05,
+      "loss": 3.8829,
+      "step": 1966592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.779275310001618e-05,
+      "loss": 3.8825,
+      "step": 1967104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.778436715250566e-05,
+      "loss": 3.871,
+      "step": 1967616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.777598120499514e-05,
+      "loss": 3.8767,
+      "step": 1968128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7767611636288354e-05,
+      "loss": 3.8662,
+      "step": 1968640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7759225688777834e-05,
+      "loss": 3.8748,
+      "step": 1969152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7750839741267314e-05,
+      "loss": 3.8774,
+      "step": 1969664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7742453793756794e-05,
+      "loss": 3.8675,
+      "step": 1970176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7734067846246274e-05,
+      "loss": 3.8646,
+      "step": 1970688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7725681898735754e-05,
+      "loss": 3.8758,
+      "step": 1971200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7717295951225234e-05,
+      "loss": 3.8755,
+      "step": 1971712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7708910003714714e-05,
+      "loss": 3.8745,
+      "step": 1972224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7700540435007926e-05,
+      "loss": 3.8599,
+      "step": 1972736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7692154487497406e-05,
+      "loss": 3.8703,
+      "step": 1973248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7683784918790615e-05,
+      "loss": 3.8689,
+      "step": 1973760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7675398971280095e-05,
+      "loss": 3.8845,
+      "step": 1974272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7667029402573308e-05,
+      "loss": 3.8802,
+      "step": 1974784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7658643455062788e-05,
+      "loss": 3.8772,
+      "step": 1975296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7650257507552268e-05,
+      "loss": 3.8637,
+      "step": 1975808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7641871560041747e-05,
+      "loss": 3.8853,
+      "step": 1976320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7633485612531227e-05,
+      "loss": 3.8775,
+      "step": 1976832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7625099665020704e-05,
+      "loss": 3.883,
+      "step": 1977344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7616713717510187e-05,
+      "loss": 3.8781,
+      "step": 1977856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7608327769999667e-05,
+      "loss": 3.8796,
+      "step": 1978368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7599941822489147e-05,
+      "loss": 3.8766,
+      "step": 1978880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7591555874978627e-05,
+      "loss": 3.8861,
+      "step": 1979392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7583186306271836e-05,
+      "loss": 3.8626,
+      "step": 1979904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7574800358761316e-05,
+      "loss": 3.8798,
+      "step": 1980416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7566414411250796e-05,
+      "loss": 3.8698,
+      "step": 1980928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.755802846374028e-05,
+      "loss": 3.874,
+      "step": 1981440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.754964251622976e-05,
+      "loss": 3.8678,
+      "step": 1981952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.754125656871924e-05,
+      "loss": 3.8692,
+      "step": 1982464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.753288700001245e-05,
+      "loss": 3.8722,
+      "step": 1982976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7524517431305658e-05,
+      "loss": 3.8746,
+      "step": 1983488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.751613148379514e-05,
+      "loss": 3.8806,
+      "step": 1984000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9819600582122803,
+      "eval_runtime": 295.4215,
+      "eval_samples_per_second": 1291.683,
+      "eval_steps_per_second": 40.366,
+      "step": 1984320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.750774553628462e-05,
+      "loss": 3.8652,
+      "step": 1984512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.74993595887741e-05,
+      "loss": 3.8698,
+      "step": 1985024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.749097364126358e-05,
+      "loss": 3.8669,
+      "step": 1985536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.748258769375306e-05,
+      "loss": 3.8799,
+      "step": 1986048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7474201746242537e-05,
+      "loss": 3.8802,
+      "step": 1986560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7465815798732017e-05,
+      "loss": 3.8664,
+      "step": 1987072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7457429851221497e-05,
+      "loss": 3.8712,
+      "step": 1987584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.744904390371098e-05,
+      "loss": 3.8665,
+      "step": 1988096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.744065795620046e-05,
+      "loss": 3.8719,
+      "step": 1988608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.743228838749367e-05,
+      "loss": 3.8841,
+      "step": 1989120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.742390243998315e-05,
+      "loss": 3.8686,
+      "step": 1989632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.741551649247263e-05,
+      "loss": 3.8727,
+      "step": 1990144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.740713054496211e-05,
+      "loss": 3.8799,
+      "step": 1990656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.739874459745159e-05,
+      "loss": 3.8586,
+      "step": 1991168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.739035864994107e-05,
+      "loss": 3.8706,
+      "step": 1991680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.738197270243055e-05,
+      "loss": 3.8585,
+      "step": 1992192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.737358675492003e-05,
+      "loss": 3.8737,
+      "step": 1992704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7365217186213242e-05,
+      "loss": 3.8765,
+      "step": 1993216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.735683123870272e-05,
+      "loss": 3.8641,
+      "step": 1993728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7348445291192202e-05,
+      "loss": 3.8813,
+      "step": 1994240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7340059343681682e-05,
+      "loss": 3.8857,
+      "step": 1994752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7331706153778623e-05,
+      "loss": 3.8756,
+      "step": 1995264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7323320206268103e-05,
+      "loss": 3.8788,
+      "step": 1995776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7314934258757583e-05,
+      "loss": 3.8776,
+      "step": 1996288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7306548311247063e-05,
+      "loss": 3.8776,
+      "step": 1996800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7298162363736543e-05,
+      "loss": 3.8682,
+      "step": 1997312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7289792795029756e-05,
+      "loss": 3.8705,
+      "step": 1997824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7281406847519236e-05,
+      "loss": 3.8772,
+      "step": 1998336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7273020900008716e-05,
+      "loss": 3.8671,
+      "step": 1998848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7264634952498192e-05,
+      "loss": 3.8722,
+      "step": 1999360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7256249004987672e-05,
+      "loss": 3.8759,
+      "step": 1999872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7247863057477155e-05,
+      "loss": 3.8791,
+      "step": 2000384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7239493488770365e-05,
+      "loss": 3.8861,
+      "step": 2000896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7231107541259845e-05,
+      "loss": 3.8747,
+      "step": 2001408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7222721593749324e-05,
+      "loss": 3.8697,
+      "step": 2001920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7214352025042537e-05,
+      "loss": 3.8739,
+      "step": 2002432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7205966077532017e-05,
+      "loss": 3.8772,
+      "step": 2002944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7197580130021497e-05,
+      "loss": 3.8655,
+      "step": 2003456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7189194182510977e-05,
+      "loss": 3.8654,
+      "step": 2003968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7180808235000457e-05,
+      "loss": 3.8661,
+      "step": 2004480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7172422287489937e-05,
+      "loss": 3.8656,
+      "step": 2004992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7164036339979417e-05,
+      "loss": 3.875,
+      "step": 2005504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7155650392468897e-05,
+      "loss": 3.8739,
+      "step": 2006016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7147264444958373e-05,
+      "loss": 3.8737,
+      "step": 2006528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7138878497447857e-05,
+      "loss": 3.8694,
+      "step": 2007040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7130492549937337e-05,
+      "loss": 3.8763,
+      "step": 2007552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7122122981230546e-05,
+      "loss": 3.8707,
+      "step": 2008064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7113737033720026e-05,
+      "loss": 3.8796,
+      "step": 2008576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7105351086209506e-05,
+      "loss": 3.8485,
+      "step": 2009088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7096965138698985e-05,
+      "loss": 3.8795,
+      "step": 2009600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7088579191188465e-05,
+      "loss": 3.87,
+      "step": 2010112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.708019324367795e-05,
+      "loss": 3.8814,
+      "step": 2010624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.707180729616743e-05,
+      "loss": 3.8661,
+      "step": 2011136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7063421348656905e-05,
+      "loss": 3.8669,
+      "step": 2011648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7055051779950118e-05,
+      "loss": 3.8646,
+      "step": 2012160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7046665832439598e-05,
+      "loss": 3.8711,
+      "step": 2012672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7038279884929078e-05,
+      "loss": 3.8777,
+      "step": 2013184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7029893937418558e-05,
+      "loss": 3.8571,
+      "step": 2013696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.702152436871177e-05,
+      "loss": 3.8823,
+      "step": 2014208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.701313842120125e-05,
+      "loss": 3.8684,
+      "step": 2014720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7004752473690727e-05,
+      "loss": 3.8476,
+      "step": 2015232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6996366526180207e-05,
+      "loss": 3.882,
+      "step": 2015744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.698799695747342e-05,
+      "loss": 3.8613,
+      "step": 2016256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.697962738876663e-05,
+      "loss": 3.8515,
+      "step": 2016768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.697124144125611e-05,
+      "loss": 3.8737,
+      "step": 2017280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.696285549374559e-05,
+      "loss": 3.8691,
+      "step": 2017792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.695446954623507e-05,
+      "loss": 3.8595,
+      "step": 2018304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.694608359872455e-05,
+      "loss": 3.8788,
+      "step": 2018816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6937714030017764e-05,
+      "loss": 3.8522,
+      "step": 2019328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6929328082507244e-05,
+      "loss": 3.8607,
+      "step": 2019840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6920942134996724e-05,
+      "loss": 3.8743,
+      "step": 2020352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.69125561874862e-05,
+      "loss": 3.8701,
+      "step": 2020864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.690417023997568e-05,
+      "loss": 3.878,
+      "step": 2021376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.689578429246516e-05,
+      "loss": 3.8717,
+      "step": 2021888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.688739834495464e-05,
+      "loss": 3.8802,
+      "step": 2022400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6879028776247853e-05,
+      "loss": 3.8631,
+      "step": 2022912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6870642828737333e-05,
+      "loss": 3.8778,
+      "step": 2023424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6862273260030545e-05,
+      "loss": 3.86,
+      "step": 2023936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6853887312520022e-05,
+      "loss": 3.8659,
+      "step": 2024448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.68455013650095e-05,
+      "loss": 3.8697,
+      "step": 2024960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.683711541749898e-05,
+      "loss": 3.8685,
+      "step": 2025472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6828729469988465e-05,
+      "loss": 3.8794,
+      "step": 2025984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6820343522477945e-05,
+      "loss": 3.8782,
+      "step": 2026496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6811957574967425e-05,
+      "loss": 3.8639,
+      "step": 2027008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6803571627456905e-05,
+      "loss": 3.8615,
+      "step": 2027520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.679518567994638e-05,
+      "loss": 3.8721,
+      "step": 2028032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.678679973243586e-05,
+      "loss": 3.8779,
+      "step": 2028544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.677841378492534e-05,
+      "loss": 3.8668,
+      "step": 2029056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.677002783741482e-05,
+      "loss": 3.8782,
+      "step": 2029568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6761674647511766e-05,
+      "loss": 3.8758,
+      "step": 2030080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6753288700001246e-05,
+      "loss": 3.8755,
+      "step": 2030592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6744902752490726e-05,
+      "loss": 3.8687,
+      "step": 2031104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6736516804980203e-05,
+      "loss": 3.8653,
+      "step": 2031616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6728130857469683e-05,
+      "loss": 3.8765,
+      "step": 2032128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6719744909959166e-05,
+      "loss": 3.8749,
+      "step": 2032640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6711358962448646e-05,
+      "loss": 3.8742,
+      "step": 2033152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6702973014938126e-05,
+      "loss": 3.8678,
+      "step": 2033664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6694619825035068e-05,
+      "loss": 3.8576,
+      "step": 2034176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6686233877524548e-05,
+      "loss": 3.8708,
+      "step": 2034688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6677847930014028e-05,
+      "loss": 3.882,
+      "step": 2035200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6669461982503507e-05,
+      "loss": 3.8701,
+      "step": 2035712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.666109241379672e-05,
+      "loss": 3.8699,
+      "step": 2036224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.66527064662862e-05,
+      "loss": 3.867,
+      "step": 2036736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6644320518775676e-05,
+      "loss": 3.8622,
+      "step": 2037248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6635934571265156e-05,
+      "loss": 3.8724,
+      "step": 2037760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6627548623754636e-05,
+      "loss": 3.8576,
+      "step": 2038272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6619179055047852e-05,
+      "loss": 3.8625,
+      "step": 2038784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.661079310753733e-05,
+      "loss": 3.871,
+      "step": 2039296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.660240716002681e-05,
+      "loss": 3.8737,
+      "step": 2039808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.659402121251629e-05,
+      "loss": 3.8662,
+      "step": 2040320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.658563526500577e-05,
+      "loss": 3.863,
+      "step": 2040832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.657724931749525e-05,
+      "loss": 3.8661,
+      "step": 2041344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.656886336998473e-05,
+      "loss": 3.8756,
+      "step": 2041856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.656049380127794e-05,
+      "loss": 3.8676,
+      "step": 2042368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.655210785376742e-05,
+      "loss": 3.8809,
+      "step": 2042880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.65437219062569e-05,
+      "loss": 3.8773,
+      "step": 2043392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.653533595874638e-05,
+      "loss": 3.8705,
+      "step": 2043904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.652696639003959e-05,
+      "loss": 3.8698,
+      "step": 2044416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6518580442529073e-05,
+      "loss": 3.8638,
+      "step": 2044928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6510194495018553e-05,
+      "loss": 3.8706,
+      "step": 2045440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6501808547508033e-05,
+      "loss": 3.875,
+      "step": 2045952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6493438978801242e-05,
+      "loss": 3.8601,
+      "step": 2046464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6485053031290722e-05,
+      "loss": 3.8601,
+      "step": 2046976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6476667083780202e-05,
+      "loss": 3.87,
+      "step": 2047488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6468281136269682e-05,
+      "loss": 3.8718,
+      "step": 2048000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6459895188759162e-05,
+      "loss": 3.8666,
+      "step": 2048512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6451509241248642e-05,
+      "loss": 3.8584,
+      "step": 2049024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6443123293738122e-05,
+      "loss": 3.8684,
+      "step": 2049536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.643475372503133e-05,
+      "loss": 3.8602,
+      "step": 2050048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.642636777752081e-05,
+      "loss": 3.8829,
+      "step": 2050560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.641798183001029e-05,
+      "loss": 3.8723,
+      "step": 2051072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6409595882499775e-05,
+      "loss": 3.8747,
+      "step": 2051584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6401209934989254e-05,
+      "loss": 3.8607,
+      "step": 2052096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6392823987478734e-05,
+      "loss": 3.8809,
+      "step": 2052608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6384438039968214e-05,
+      "loss": 3.871,
+      "step": 2053120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.637605209245769e-05,
+      "loss": 3.879,
+      "step": 2053632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6367682523750903e-05,
+      "loss": 3.8788,
+      "step": 2054144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6359312955044116e-05,
+      "loss": 3.8741,
+      "step": 2054656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6350927007533596e-05,
+      "loss": 3.8704,
+      "step": 2055168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6342541060023076e-05,
+      "loss": 3.8833,
+      "step": 2055680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6334155112512556e-05,
+      "loss": 3.857,
+      "step": 2056192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6325769165002036e-05,
+      "loss": 3.8737,
+      "step": 2056704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6317383217491512e-05,
+      "loss": 3.8656,
+      "step": 2057216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6308997269980996e-05,
+      "loss": 3.8699,
+      "step": 2057728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6300611322470476e-05,
+      "loss": 3.8622,
+      "step": 2058240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6292225374959956e-05,
+      "loss": 3.8646,
+      "step": 2058752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6283855806253165e-05,
+      "loss": 3.8651,
+      "step": 2059264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6275486237546377e-05,
+      "loss": 3.8747,
+      "step": 2059776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.626711666883959e-05,
+      "loss": 3.8744,
+      "step": 2060288
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9804866313934326,
+      "eval_runtime": 298.3013,
+      "eval_samples_per_second": 1279.213,
+      "eval_steps_per_second": 39.976,
+      "step": 2060640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.625873072132907e-05,
+      "loss": 3.863,
+      "step": 2060800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.625034477381855e-05,
+      "loss": 3.864,
+      "step": 2061312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.624195882630803e-05,
+      "loss": 3.8621,
+      "step": 2061824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.623357287879751e-05,
+      "loss": 3.8761,
+      "step": 2062336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6225186931286986e-05,
+      "loss": 3.8731,
+      "step": 2062848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6216800983776466e-05,
+      "loss": 3.863,
+      "step": 2063360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6208415036265946e-05,
+      "loss": 3.8679,
+      "step": 2063872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.620002908875543e-05,
+      "loss": 3.8649,
+      "step": 2064384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.619164314124491e-05,
+      "loss": 3.8685,
+      "step": 2064896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.618325719373439e-05,
+      "loss": 3.8746,
+      "step": 2065408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.617487124622387e-05,
+      "loss": 3.8697,
+      "step": 2065920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6166485298713346e-05,
+      "loss": 3.8684,
+      "step": 2066432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6158099351202826e-05,
+      "loss": 3.8773,
+      "step": 2066944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6149713403692306e-05,
+      "loss": 3.8511,
+      "step": 2067456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.614132745618179e-05,
+      "loss": 3.8666,
+      "step": 2067968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.613294150867127e-05,
+      "loss": 3.8521,
+      "step": 2068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.612455556116075e-05,
+      "loss": 3.8725,
+      "step": 2068992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6116169613650225e-05,
+      "loss": 3.8767,
+      "step": 2069504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6107783666139705e-05,
+      "loss": 3.8541,
+      "step": 2070016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6099397718629185e-05,
+      "loss": 3.8758,
+      "step": 2070528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6091028149922398e-05,
+      "loss": 3.8822,
+      "step": 2071040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6082642202411878e-05,
+      "loss": 3.8713,
+      "step": 2071552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6074256254901358e-05,
+      "loss": 3.8767,
+      "step": 2072064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6065870307390838e-05,
+      "loss": 3.8736,
+      "step": 2072576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.605750073868405e-05,
+      "loss": 3.8724,
+      "step": 2073088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6049114791173527e-05,
+      "loss": 3.8657,
+      "step": 2073600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6040728843663007e-05,
+      "loss": 3.8623,
+      "step": 2074112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.603234289615249e-05,
+      "loss": 3.8682,
+      "step": 2074624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.60239733274457e-05,
+      "loss": 3.8633,
+      "step": 2075136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.601560375873891e-05,
+      "loss": 3.8666,
+      "step": 2075648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.600721781122839e-05,
+      "loss": 3.8729,
+      "step": 2076160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.599883186371787e-05,
+      "loss": 3.8772,
+      "step": 2076672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.599044591620735e-05,
+      "loss": 3.8716,
+      "step": 2077184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.598205996869683e-05,
+      "loss": 3.8767,
+      "step": 2077696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.597367402118631e-05,
+      "loss": 3.8615,
+      "step": 2078208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5965320831283253e-05,
+      "loss": 3.8728,
+      "step": 2078720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5956934883772733e-05,
+      "loss": 3.8756,
+      "step": 2079232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5948548936262213e-05,
+      "loss": 3.8537,
+      "step": 2079744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5940162988751693e-05,
+      "loss": 3.8634,
+      "step": 2080256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5931777041241173e-05,
+      "loss": 3.8634,
+      "step": 2080768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5923391093730653e-05,
+      "loss": 3.8618,
+      "step": 2081280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5915005146220133e-05,
+      "loss": 3.8688,
+      "step": 2081792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5906619198709613e-05,
+      "loss": 3.8698,
+      "step": 2082304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5898233251199093e-05,
+      "loss": 3.8677,
+      "step": 2082816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5889847303688573e-05,
+      "loss": 3.8701,
+      "step": 2083328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5881461356178053e-05,
+      "loss": 3.8704,
+      "step": 2083840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5873091787471265e-05,
+      "loss": 3.8634,
+      "step": 2084352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5864705839960745e-05,
+      "loss": 3.874,
+      "step": 2084864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5856319892450225e-05,
+      "loss": 3.8488,
+      "step": 2085376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5847933944939705e-05,
+      "loss": 3.874,
+      "step": 2085888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5839564376232914e-05,
+      "loss": 3.862,
+      "step": 2086400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5831178428722397e-05,
+      "loss": 3.8781,
+      "step": 2086912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5822792481211877e-05,
+      "loss": 3.862,
+      "step": 2087424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5814406533701354e-05,
+      "loss": 3.8595,
+      "step": 2087936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5806020586190834e-05,
+      "loss": 3.8612,
+      "step": 2088448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5797634638680314e-05,
+      "loss": 3.8649,
+      "step": 2088960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5789265069973526e-05,
+      "loss": 3.8719,
+      "step": 2089472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5780879122463006e-05,
+      "loss": 3.8511,
+      "step": 2089984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5772493174952486e-05,
+      "loss": 3.8772,
+      "step": 2090496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5764107227441966e-05,
+      "loss": 3.8688,
+      "step": 2091008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5755721279931446e-05,
+      "loss": 3.8427,
+      "step": 2091520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5747335332420926e-05,
+      "loss": 3.8764,
+      "step": 2092032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5738949384910406e-05,
+      "loss": 3.8551,
+      "step": 2092544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5730563437399886e-05,
+      "loss": 3.8485,
+      "step": 2093056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.57221938686931e-05,
+      "loss": 3.8648,
+      "step": 2093568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.571380792118258e-05,
+      "loss": 3.8653,
+      "step": 2094080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.570542197367206e-05,
+      "loss": 3.8574,
+      "step": 2094592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5697036026161535e-05,
+      "loss": 3.8743,
+      "step": 2095104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5688650078651015e-05,
+      "loss": 3.8456,
+      "step": 2095616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5680280509944227e-05,
+      "loss": 3.8554,
+      "step": 2096128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5671894562433707e-05,
+      "loss": 3.8709,
+      "step": 2096640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5663508614923187e-05,
+      "loss": 3.8698,
+      "step": 2097152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.56551390462164e-05,
+      "loss": 3.8708,
+      "step": 2097664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.564675309870588e-05,
+      "loss": 3.87,
+      "step": 2098176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5638367151195356e-05,
+      "loss": 3.8743,
+      "step": 2098688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5629981203684836e-05,
+      "loss": 3.8588,
+      "step": 2099200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5621611634978052e-05,
+      "loss": 3.8749,
+      "step": 2099712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5613225687467532e-05,
+      "loss": 3.8569,
+      "step": 2100224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.560483973995701e-05,
+      "loss": 3.8593,
+      "step": 2100736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.559645379244649e-05,
+      "loss": 3.8624,
+      "step": 2101248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.558806784493597e-05,
+      "loss": 3.8662,
+      "step": 2101760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.557968189742545e-05,
+      "loss": 3.8743,
+      "step": 2102272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.557129594991493e-05,
+      "loss": 3.8754,
+      "step": 2102784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.556291000240441e-05,
+      "loss": 3.8619,
+      "step": 2103296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5554524054893892e-05,
+      "loss": 3.8542,
+      "step": 2103808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.55461544861871e-05,
+      "loss": 3.8675,
+      "step": 2104320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.553776853867658e-05,
+      "loss": 3.873,
+      "step": 2104832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.552938259116606e-05,
+      "loss": 3.8625,
+      "step": 2105344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5520996643655537e-05,
+      "loss": 3.8737,
+      "step": 2105856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5512627074948753e-05,
+      "loss": 3.8721,
+      "step": 2106368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5504241127438233e-05,
+      "loss": 3.8692,
+      "step": 2106880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5495855179927713e-05,
+      "loss": 3.8642,
+      "step": 2107392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.548746923241719e-05,
+      "loss": 3.8577,
+      "step": 2107904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.547908328490667e-05,
+      "loss": 3.8745,
+      "step": 2108416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.547069733739615e-05,
+      "loss": 3.8731,
+      "step": 2108928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.546231138988563e-05,
+      "loss": 3.8676,
+      "step": 2109440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.545392544237511e-05,
+      "loss": 3.8661,
+      "step": 2109952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5445555873668322e-05,
+      "loss": 3.8521,
+      "step": 2110464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5437169926157802e-05,
+      "loss": 3.8664,
+      "step": 2110976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5428783978647282e-05,
+      "loss": 3.8772,
+      "step": 2111488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5420398031136762e-05,
+      "loss": 3.864,
+      "step": 2112000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5412012083626242e-05,
+      "loss": 3.8655,
+      "step": 2112512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5403642514919454e-05,
+      "loss": 3.8631,
+      "step": 2113024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5395256567408934e-05,
+      "loss": 3.861,
+      "step": 2113536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5386870619898414e-05,
+      "loss": 3.8654,
+      "step": 2114048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5378484672387894e-05,
+      "loss": 3.8603,
+      "step": 2114560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5370115103681103e-05,
+      "loss": 3.8582,
+      "step": 2115072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5361729156170583e-05,
+      "loss": 3.866,
+      "step": 2115584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5353343208660063e-05,
+      "loss": 3.8682,
+      "step": 2116096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5344957261149547e-05,
+      "loss": 3.8623,
+      "step": 2116608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5336587692442756e-05,
+      "loss": 3.8624,
+      "step": 2117120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5328201744932236e-05,
+      "loss": 3.8596,
+      "step": 2117632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5319815797421716e-05,
+      "loss": 3.8716,
+      "step": 2118144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5311429849911192e-05,
+      "loss": 3.8633,
+      "step": 2118656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5303060281204408e-05,
+      "loss": 3.8755,
+      "step": 2119168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5294674333693888e-05,
+      "loss": 3.87,
+      "step": 2119680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5286288386183368e-05,
+      "loss": 3.8655,
+      "step": 2120192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5277918817476577e-05,
+      "loss": 3.8685,
+      "step": 2120704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5269532869966057e-05,
+      "loss": 3.8621,
+      "step": 2121216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5261146922455537e-05,
+      "loss": 3.8653,
+      "step": 2121728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5252760974945019e-05,
+      "loss": 3.8701,
+      "step": 2122240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5244375027434497e-05,
+      "loss": 3.858,
+      "step": 2122752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5235989079923977e-05,
+      "loss": 3.8581,
+      "step": 2123264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5227603132413457e-05,
+      "loss": 3.866,
+      "step": 2123776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5219217184902937e-05,
+      "loss": 3.8652,
+      "step": 2124288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5210847616196147e-05,
+      "loss": 3.86,
+      "step": 2124800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5202461668685627e-05,
+      "loss": 3.8551,
+      "step": 2125312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.519409209997884e-05,
+      "loss": 3.8599,
+      "step": 2125824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5185706152468318e-05,
+      "loss": 3.86,
+      "step": 2126336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5177320204957798e-05,
+      "loss": 3.8797,
+      "step": 2126848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5168934257447278e-05,
+      "loss": 3.8648,
+      "step": 2127360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.516054830993676e-05,
+      "loss": 3.8716,
+      "step": 2127872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.515216236242624e-05,
+      "loss": 3.858,
+      "step": 2128384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.514377641491572e-05,
+      "loss": 3.8696,
+      "step": 2128896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.51353904674052e-05,
+      "loss": 3.8679,
+      "step": 2129408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.512702089869841e-05,
+      "loss": 3.8724,
+      "step": 2129920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.511863495118789e-05,
+      "loss": 3.8775,
+      "step": 2130432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.511024900367737e-05,
+      "loss": 3.8663,
+      "step": 2130944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5101863056166849e-05,
+      "loss": 3.8694,
+      "step": 2131456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5093477108656329e-05,
+      "loss": 3.871,
+      "step": 2131968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5085091161145808e-05,
+      "loss": 3.8584,
+      "step": 2132480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.507670521363529e-05,
+      "loss": 3.8681,
+      "step": 2132992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.506831926612477e-05,
+      "loss": 3.8632,
+      "step": 2133504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.505994969741798e-05,
+      "loss": 3.8635,
+      "step": 2134016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.505156374990746e-05,
+      "loss": 3.8563,
+      "step": 2134528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.504319418120067e-05,
+      "loss": 3.8631,
+      "step": 2135040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5034824612493884e-05,
+      "loss": 3.862,
+      "step": 2135552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5026438664983364e-05,
+      "loss": 3.8649,
+      "step": 2136064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5018052717472844e-05,
+      "loss": 3.8708,
+      "step": 2136576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9788742065429688,
+      "eval_runtime": 297.8275,
+      "eval_samples_per_second": 1281.248,
+      "eval_steps_per_second": 40.04,
+      "step": 2136960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5009666769962322e-05,
+      "loss": 3.8641,
+      "step": 2137088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5001280822451802e-05,
+      "loss": 3.8561,
+      "step": 2137600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4992894874941282e-05,
+      "loss": 3.8579,
+      "step": 2138112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4984508927430762e-05,
+      "loss": 3.8702,
+      "step": 2138624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4976122979920242e-05,
+      "loss": 3.8666,
+      "step": 2139136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4967737032409724e-05,
+      "loss": 3.8615,
+      "step": 2139648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4959351084899204e-05,
+      "loss": 3.8679,
+      "step": 2140160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.495096513738868e-05,
+      "loss": 3.8579,
+      "step": 2140672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4942579189878162e-05,
+      "loss": 3.8613,
+      "step": 2141184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4934209621171374e-05,
+      "loss": 3.8663,
+      "step": 2141696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4925823673660853e-05,
+      "loss": 3.8708,
+      "step": 2142208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4917454104954065e-05,
+      "loss": 3.8623,
+      "step": 2142720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4909068157443545e-05,
+      "loss": 3.8734,
+      "step": 2143232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4900698588736756e-05,
+      "loss": 3.8447,
+      "step": 2143744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4892312641226236e-05,
+      "loss": 3.8639,
+      "step": 2144256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4883926693715716e-05,
+      "loss": 3.8496,
+      "step": 2144768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4875540746205196e-05,
+      "loss": 3.8669,
+      "step": 2145280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4867154798694677e-05,
+      "loss": 3.8706,
+      "step": 2145792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4858768851184154e-05,
+      "loss": 3.8517,
+      "step": 2146304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4850382903673634e-05,
+      "loss": 3.8682,
+      "step": 2146816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4841996956163116e-05,
+      "loss": 3.879,
+      "step": 2147328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4833627387456325e-05,
+      "loss": 3.8644,
+      "step": 2147840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4825241439945806e-05,
+      "loss": 3.8708,
+      "step": 2148352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4816855492435286e-05,
+      "loss": 3.8705,
+      "step": 2148864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4808469544924766e-05,
+      "loss": 3.8689,
+      "step": 2149376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4800083597414246e-05,
+      "loss": 3.864,
+      "step": 2149888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4791697649903726e-05,
+      "loss": 3.8576,
+      "step": 2150400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4783311702393208e-05,
+      "loss": 3.8657,
+      "step": 2150912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4774925754882684e-05,
+      "loss": 3.8604,
+      "step": 2151424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4766539807372164e-05,
+      "loss": 3.8618,
+      "step": 2151936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4758170238665379e-05,
+      "loss": 3.8708,
+      "step": 2152448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4749784291154858e-05,
+      "loss": 3.8699,
+      "step": 2152960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4741398343644337e-05,
+      "loss": 3.8674,
+      "step": 2153472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4733012396133817e-05,
+      "loss": 3.8755,
+      "step": 2153984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.472464282742703e-05,
+      "loss": 3.8599,
+      "step": 2154496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.471627325872024e-05,
+      "loss": 3.8681,
+      "step": 2155008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.470788731120972e-05,
+      "loss": 3.8661,
+      "step": 2155520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.46995013636992e-05,
+      "loss": 3.8511,
+      "step": 2156032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.469111541618868e-05,
+      "loss": 3.8611,
+      "step": 2156544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4682729468678158e-05,
+      "loss": 3.8601,
+      "step": 2157056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4674343521167638e-05,
+      "loss": 3.8559,
+      "step": 2157568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4665957573657118e-05,
+      "loss": 3.8641,
+      "step": 2158080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.46575716261466e-05,
+      "loss": 3.8675,
+      "step": 2158592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4649202057439809e-05,
+      "loss": 3.863,
+      "step": 2159104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4640816109929289e-05,
+      "loss": 3.8656,
+      "step": 2159616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.463243016241877e-05,
+      "loss": 3.8692,
+      "step": 2160128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.462404421490825e-05,
+      "loss": 3.8558,
+      "step": 2160640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4615674646201461e-05,
+      "loss": 3.8693,
+      "step": 2161152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4607288698690941e-05,
+      "loss": 3.8476,
+      "step": 2161664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4598902751180421e-05,
+      "loss": 3.8657,
+      "step": 2162176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4590516803669901e-05,
+      "loss": 3.8593,
+      "step": 2162688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4582147234963112e-05,
+      "loss": 3.8746,
+      "step": 2163200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4573761287452592e-05,
+      "loss": 3.8592,
+      "step": 2163712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4565375339942072e-05,
+      "loss": 3.8539,
+      "step": 2164224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4557005771235282e-05,
+      "loss": 3.8592,
+      "step": 2164736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4548619823724762e-05,
+      "loss": 3.8617,
+      "step": 2165248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4540233876214242e-05,
+      "loss": 3.8702,
+      "step": 2165760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4531847928703724e-05,
+      "loss": 3.8427,
+      "step": 2166272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4523461981193204e-05,
+      "loss": 3.8758,
+      "step": 2166784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4515092412486415e-05,
+      "loss": 3.863,
+      "step": 2167296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4506706464975895e-05,
+      "loss": 3.8402,
+      "step": 2167808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4498320517465375e-05,
+      "loss": 3.8705,
+      "step": 2168320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4489934569954855e-05,
+      "loss": 3.8495,
+      "step": 2168832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4481548622444335e-05,
+      "loss": 3.8462,
+      "step": 2169344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4473162674933813e-05,
+      "loss": 3.8569,
+      "step": 2169856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4464793106227025e-05,
+      "loss": 3.865,
+      "step": 2170368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4456407158716507e-05,
+      "loss": 3.8598,
+      "step": 2170880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4448021211205984e-05,
+      "loss": 3.8629,
+      "step": 2171392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4439635263695464e-05,
+      "loss": 3.8469,
+      "step": 2171904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4431265694988678e-05,
+      "loss": 3.8488,
+      "step": 2172416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4422879747478158e-05,
+      "loss": 3.8655,
+      "step": 2172928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4414493799967634e-05,
+      "loss": 3.8691,
+      "step": 2173440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4406107852457116e-05,
+      "loss": 3.8671,
+      "step": 2173952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4397721904946596e-05,
+      "loss": 3.8651,
+      "step": 2174464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4389335957436076e-05,
+      "loss": 3.8694,
+      "step": 2174976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4380950009925556e-05,
+      "loss": 3.8566,
+      "step": 2175488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4372580441218767e-05,
+      "loss": 3.8652,
+      "step": 2176000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4364194493708246e-05,
+      "loss": 3.8558,
+      "step": 2176512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4355808546197726e-05,
+      "loss": 3.8567,
+      "step": 2177024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4347422598687208e-05,
+      "loss": 3.8613,
+      "step": 2177536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4339036651176688e-05,
+      "loss": 3.853,
+      "step": 2178048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4330650703666165e-05,
+      "loss": 3.871,
+      "step": 2178560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4322264756155646e-05,
+      "loss": 3.8711,
+      "step": 2179072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4313878808645126e-05,
+      "loss": 3.8613,
+      "step": 2179584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4305509239938339e-05,
+      "loss": 3.8435,
+      "step": 2180096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4297123292427817e-05,
+      "loss": 3.8636,
+      "step": 2180608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4288737344917297e-05,
+      "loss": 3.8671,
+      "step": 2181120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4280351397406777e-05,
+      "loss": 3.8615,
+      "step": 2181632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4271965449896257e-05,
+      "loss": 3.8687,
+      "step": 2182144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.42636122599932e-05,
+      "loss": 3.872,
+      "step": 2182656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.425522631248268e-05,
+      "loss": 3.8611,
+      "step": 2183168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4246840364972162e-05,
+      "loss": 3.8631,
+      "step": 2183680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4238454417461638e-05,
+      "loss": 3.8557,
+      "step": 2184192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4230068469951118e-05,
+      "loss": 3.8705,
+      "step": 2184704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.42216825224406e-05,
+      "loss": 3.8648,
+      "step": 2185216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4213312953733812e-05,
+      "loss": 3.867,
+      "step": 2185728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4204927006223289e-05,
+      "loss": 3.8595,
+      "step": 2186240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.419654105871277e-05,
+      "loss": 3.8501,
+      "step": 2186752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.418815511120225e-05,
+      "loss": 3.8634,
+      "step": 2187264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.417976916369173e-05,
+      "loss": 3.8737,
+      "step": 2187776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.417138321618121e-05,
+      "loss": 3.8644,
+      "step": 2188288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4162997268670692e-05,
+      "loss": 3.8542,
+      "step": 2188800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4154611321160172e-05,
+      "loss": 3.8627,
+      "step": 2189312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4146241752453381e-05,
+      "loss": 3.855,
+      "step": 2189824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4137855804942863e-05,
+      "loss": 3.8607,
+      "step": 2190336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4129469857432343e-05,
+      "loss": 3.8588,
+      "step": 2190848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4121100288725554e-05,
+      "loss": 3.8534,
+      "step": 2191360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4112714341215034e-05,
+      "loss": 3.8593,
+      "step": 2191872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4104328393704513e-05,
+      "loss": 3.8646,
+      "step": 2192384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4095942446193993e-05,
+      "loss": 3.8601,
+      "step": 2192896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4087572877487204e-05,
+      "loss": 3.8543,
+      "step": 2193408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4079186929976684e-05,
+      "loss": 3.8581,
+      "step": 2193920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4070800982466164e-05,
+      "loss": 3.8663,
+      "step": 2194432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4062431413759375e-05,
+      "loss": 3.8625,
+      "step": 2194944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4054045466248855e-05,
+      "loss": 3.8679,
+      "step": 2195456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4045659518738335e-05,
+      "loss": 3.8642,
+      "step": 2195968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4037273571227816e-05,
+      "loss": 3.8673,
+      "step": 2196480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4028904002521026e-05,
+      "loss": 3.8633,
+      "step": 2196992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4020518055010507e-05,
+      "loss": 3.8553,
+      "step": 2197504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4012132107499987e-05,
+      "loss": 3.862,
+      "step": 2198016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4003746159989467e-05,
+      "loss": 3.864,
+      "step": 2198528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3995360212478945e-05,
+      "loss": 3.8543,
+      "step": 2199040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3986974264968425e-05,
+      "loss": 3.8592,
+      "step": 2199552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3978588317457905e-05,
+      "loss": 3.8572,
+      "step": 2200064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3970202369947385e-05,
+      "loss": 3.8618,
+      "step": 2200576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3961816422436865e-05,
+      "loss": 3.8601,
+      "step": 2201088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3953430474926347e-05,
+      "loss": 3.8518,
+      "step": 2201600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3945044527415827e-05,
+      "loss": 3.8528,
+      "step": 2202112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3936674958709036e-05,
+      "loss": 3.8554,
+      "step": 2202624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3928289011198518e-05,
+      "loss": 3.8744,
+      "step": 2203136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3919903063687998e-05,
+      "loss": 3.8638,
+      "step": 2203648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3911517116177474e-05,
+      "loss": 3.8643,
+      "step": 2204160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3903131168666956e-05,
+      "loss": 3.8551,
+      "step": 2204672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3894745221156436e-05,
+      "loss": 3.8674,
+      "step": 2205184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3886375652449648e-05,
+      "loss": 3.8656,
+      "step": 2205696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3877989704939126e-05,
+      "loss": 3.8663,
+      "step": 2206208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3869603757428606e-05,
+      "loss": 3.8716,
+      "step": 2206720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3861217809918086e-05,
+      "loss": 3.862,
+      "step": 2207232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3852831862407566e-05,
+      "loss": 3.8695,
+      "step": 2207744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3844445914897048e-05,
+      "loss": 3.8692,
+      "step": 2208256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3836059967386528e-05,
+      "loss": 3.853,
+      "step": 2208768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3827674019876005e-05,
+      "loss": 3.8647,
+      "step": 2209280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3819288072365486e-05,
+      "loss": 3.8587,
+      "step": 2209792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3810918503658699e-05,
+      "loss": 3.8625,
+      "step": 2210304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3802532556148179e-05,
+      "loss": 3.8541,
+      "step": 2210816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3794146608637657e-05,
+      "loss": 3.8594,
+      "step": 2211328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3785760661127137e-05,
+      "loss": 3.8565,
+      "step": 2211840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3777374713616617e-05,
+      "loss": 3.8604,
+      "step": 2212352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.376900514490983e-05,
+      "loss": 3.8653,
+      "step": 2212864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.977689266204834,
+      "eval_runtime": 297.5935,
+      "eval_samples_per_second": 1282.256,
+      "eval_steps_per_second": 40.071,
+      "step": 2213280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3760619197399308e-05,
+      "loss": 3.8737,
+      "step": 2213376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3752233249888787e-05,
+      "loss": 3.8497,
+      "step": 2213888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3743847302378267e-05,
+      "loss": 3.8558,
+      "step": 2214400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3735461354867749e-05,
+      "loss": 3.8655,
+      "step": 2214912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3727075407357229e-05,
+      "loss": 3.8599,
+      "step": 2215424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3718689459846709e-05,
+      "loss": 3.8592,
+      "step": 2215936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3710303512336187e-05,
+      "loss": 3.8616,
+      "step": 2216448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3701917564825667e-05,
+      "loss": 3.8549,
+      "step": 2216960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3693531617315147e-05,
+      "loss": 3.8554,
+      "step": 2217472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3685145669804627e-05,
+      "loss": 3.8671,
+      "step": 2217984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3676759722294109e-05,
+      "loss": 3.8634,
+      "step": 2218496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3668373774783589e-05,
+      "loss": 3.8585,
+      "step": 2219008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3659987827273069e-05,
+      "loss": 3.8724,
+      "step": 2219520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3651601879762547e-05,
+      "loss": 3.8386,
+      "step": 2220032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3643215932252027e-05,
+      "loss": 3.8572,
+      "step": 2220544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3634829984741507e-05,
+      "loss": 3.8465,
+      "step": 2221056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3626444037230987e-05,
+      "loss": 3.8634,
+      "step": 2221568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3618058089720467e-05,
+      "loss": 3.8631,
+      "step": 2222080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3609672142209948e-05,
+      "loss": 3.8518,
+      "step": 2222592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3601286194699428e-05,
+      "loss": 3.8579,
+      "step": 2223104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3592916625992637e-05,
+      "loss": 3.8795,
+      "step": 2223616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3584547057285848e-05,
+      "loss": 3.8597,
+      "step": 2224128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3576161109775328e-05,
+      "loss": 3.8665,
+      "step": 2224640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.356777516226481e-05,
+      "loss": 3.8686,
+      "step": 2225152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.355938921475429e-05,
+      "loss": 3.8667,
+      "step": 2225664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.355100326724377e-05,
+      "loss": 3.8586,
+      "step": 2226176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.354261731973325e-05,
+      "loss": 3.8549,
+      "step": 2226688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3534231372222728e-05,
+      "loss": 3.8608,
+      "step": 2227200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.352586180351594e-05,
+      "loss": 3.8553,
+      "step": 2227712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.351747585600542e-05,
+      "loss": 3.8567,
+      "step": 2228224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3509089908494902e-05,
+      "loss": 3.8666,
+      "step": 2228736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3500703960984379e-05,
+      "loss": 3.8671,
+      "step": 2229248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3492318013473859e-05,
+      "loss": 3.8638,
+      "step": 2229760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.348393206596334e-05,
+      "loss": 3.8694,
+      "step": 2230272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.347554611845282e-05,
+      "loss": 3.8559,
+      "step": 2230784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.346717654974603e-05,
+      "loss": 3.8656,
+      "step": 2231296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3458790602235511e-05,
+      "loss": 3.8647,
+      "step": 2231808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3450404654724991e-05,
+      "loss": 3.8444,
+      "step": 2232320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3442018707214471e-05,
+      "loss": 3.8615,
+      "step": 2232832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.343363275970395e-05,
+      "loss": 3.8501,
+      "step": 2233344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.342524681219343e-05,
+      "loss": 3.8486,
+      "step": 2233856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3416860864682909e-05,
+      "loss": 3.8603,
+      "step": 2234368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3408474917172389e-05,
+      "loss": 3.8707,
+      "step": 2234880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3400105348465603e-05,
+      "loss": 3.8539,
+      "step": 2235392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3391719400955083e-05,
+      "loss": 3.8611,
+      "step": 2235904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.338333345344456e-05,
+      "loss": 3.8646,
+      "step": 2236416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3374947505934041e-05,
+      "loss": 3.8535,
+      "step": 2236928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3366577937227254e-05,
+      "loss": 3.8655,
+      "step": 2237440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3358191989716732e-05,
+      "loss": 3.846,
+      "step": 2237952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3349806042206212e-05,
+      "loss": 3.8587,
+      "step": 2238464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3341436473499425e-05,
+      "loss": 3.8583,
+      "step": 2238976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3333050525988904e-05,
+      "loss": 3.8677,
+      "step": 2239488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3324664578478383e-05,
+      "loss": 3.8566,
+      "step": 2240000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3316278630967863e-05,
+      "loss": 3.8489,
+      "step": 2240512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3307892683457343e-05,
+      "loss": 3.8536,
+      "step": 2241024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3299506735946823e-05,
+      "loss": 3.8577,
+      "step": 2241536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3291137167240033e-05,
+      "loss": 3.8673,
+      "step": 2242048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3282751219729513e-05,
+      "loss": 3.8437,
+      "step": 2242560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3274365272218995e-05,
+      "loss": 3.8662,
+      "step": 2243072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3265979324708475e-05,
+      "loss": 3.8577,
+      "step": 2243584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3257593377197955e-05,
+      "loss": 3.843,
+      "step": 2244096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3249223808491166e-05,
+      "loss": 3.8632,
+      "step": 2244608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3240837860980646e-05,
+      "loss": 3.8488,
+      "step": 2245120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3232451913470126e-05,
+      "loss": 3.8414,
+      "step": 2245632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3224065965959606e-05,
+      "loss": 3.8495,
+      "step": 2246144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3215696397252816e-05,
+      "loss": 3.8613,
+      "step": 2246656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3207310449742296e-05,
+      "loss": 3.8566,
+      "step": 2247168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3198924502231776e-05,
+      "loss": 3.8513,
+      "step": 2247680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3190554933524987e-05,
+      "loss": 3.848,
+      "step": 2248192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3182168986014467e-05,
+      "loss": 3.8472,
+      "step": 2248704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3173783038503949e-05,
+      "loss": 3.8577,
+      "step": 2249216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3165397090993429e-05,
+      "loss": 3.8639,
+      "step": 2249728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3157011143482909e-05,
+      "loss": 3.8622,
+      "step": 2250240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3148625195972387e-05,
+      "loss": 3.8659,
+      "step": 2250752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3140239248461867e-05,
+      "loss": 3.8621,
+      "step": 2251264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3131853300951347e-05,
+      "loss": 3.8538,
+      "step": 2251776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.312348373224456e-05,
+      "loss": 3.8599,
+      "step": 2252288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3115097784734038e-05,
+      "loss": 3.8533,
+      "step": 2252800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3106711837223517e-05,
+      "loss": 3.8564,
+      "step": 2253312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3098325889712997e-05,
+      "loss": 3.85,
+      "step": 2253824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3089939942202477e-05,
+      "loss": 3.8532,
+      "step": 2254336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3081553994691959e-05,
+      "loss": 3.8694,
+      "step": 2254848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3073168047181439e-05,
+      "loss": 3.8662,
+      "step": 2255360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3064782099670916e-05,
+      "loss": 3.8609,
+      "step": 2255872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3056396152160397e-05,
+      "loss": 3.8426,
+      "step": 2256384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.304804296225734e-05,
+      "loss": 3.859,
+      "step": 2256896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.303965701474682e-05,
+      "loss": 3.8632,
+      "step": 2257408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.30312710672363e-05,
+      "loss": 3.8582,
+      "step": 2257920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.302288511972578e-05,
+      "loss": 3.8606,
+      "step": 2258432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.301449917221526e-05,
+      "loss": 3.8703,
+      "step": 2258944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3006113224704742e-05,
+      "loss": 3.8576,
+      "step": 2259456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2997727277194219e-05,
+      "loss": 3.8564,
+      "step": 2259968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2989341329683699e-05,
+      "loss": 3.8532,
+      "step": 2260480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.298095538217318e-05,
+      "loss": 3.8638,
+      "step": 2260992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.297258581346639e-05,
+      "loss": 3.8631,
+      "step": 2261504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.296419986595587e-05,
+      "loss": 3.8649,
+      "step": 2262016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2955813918445351e-05,
+      "loss": 3.8524,
+      "step": 2262528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.294742797093483e-05,
+      "loss": 3.8489,
+      "step": 2263040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.293904202342431e-05,
+      "loss": 3.8549,
+      "step": 2263552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.293065607591379e-05,
+      "loss": 3.8676,
+      "step": 2264064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2922270128403272e-05,
+      "loss": 3.8608,
+      "step": 2264576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2913884180892749e-05,
+      "loss": 3.8509,
+      "step": 2265088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2905514612185961e-05,
+      "loss": 3.8582,
+      "step": 2265600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2897128664675443e-05,
+      "loss": 3.8495,
+      "step": 2266112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2888742717164923e-05,
+      "loss": 3.8574,
+      "step": 2266624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.28803567696544e-05,
+      "loss": 3.8527,
+      "step": 2267136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2872003579751343e-05,
+      "loss": 3.8516,
+      "step": 2267648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2863617632240823e-05,
+      "loss": 3.8581,
+      "step": 2268160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2855231684730305e-05,
+      "loss": 3.8602,
+      "step": 2268672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2846845737219784e-05,
+      "loss": 3.853,
+      "step": 2269184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2838459789709264e-05,
+      "loss": 3.8532,
+      "step": 2269696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2830073842198744e-05,
+      "loss": 3.8528,
+      "step": 2270208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2821704273491955e-05,
+      "loss": 3.864,
+      "step": 2270720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2813318325981435e-05,
+      "loss": 3.8572,
+      "step": 2271232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2804932378470915e-05,
+      "loss": 3.8652,
+      "step": 2271744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2796546430960397e-05,
+      "loss": 3.8577,
+      "step": 2272256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2788160483449873e-05,
+      "loss": 3.8651,
+      "step": 2272768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2779790914743087e-05,
+      "loss": 3.8582,
+      "step": 2273280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2771404967232567e-05,
+      "loss": 3.8504,
+      "step": 2273792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2763019019722044e-05,
+      "loss": 3.862,
+      "step": 2274304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2754633072211526e-05,
+      "loss": 3.8563,
+      "step": 2274816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2746247124701006e-05,
+      "loss": 3.85,
+      "step": 2275328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2737861177190486e-05,
+      "loss": 3.8543,
+      "step": 2275840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2729491608483696e-05,
+      "loss": 3.8555,
+      "step": 2276352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2721105660973176e-05,
+      "loss": 3.8583,
+      "step": 2276864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2712719713462656e-05,
+      "loss": 3.8499,
+      "step": 2277376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2704333765952136e-05,
+      "loss": 3.8504,
+      "step": 2277888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2695947818441616e-05,
+      "loss": 3.8512,
+      "step": 2278400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2687561870931098e-05,
+      "loss": 3.8522,
+      "step": 2278912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2679192302224307e-05,
+      "loss": 3.8691,
+      "step": 2279424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2670806354713789e-05,
+      "loss": 3.8607,
+      "step": 2279936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2662420407203269e-05,
+      "loss": 3.8635,
+      "step": 2280448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2654034459692748e-05,
+      "loss": 3.851,
+      "step": 2280960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2645648512182227e-05,
+      "loss": 3.8639,
+      "step": 2281472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2637262564671707e-05,
+      "loss": 3.8614,
+      "step": 2281984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2628876617161187e-05,
+      "loss": 3.8639,
+      "step": 2282496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2620507048454399e-05,
+      "loss": 3.8615,
+      "step": 2283008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2612121100943877e-05,
+      "loss": 3.8593,
+      "step": 2283520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2603735153433357e-05,
+      "loss": 3.8677,
+      "step": 2284032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.259536558472657e-05,
+      "loss": 3.8654,
+      "step": 2284544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2586979637216048e-05,
+      "loss": 3.8513,
+      "step": 2285056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2578593689705528e-05,
+      "loss": 3.8572,
+      "step": 2285568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2570207742195008e-05,
+      "loss": 3.8578,
+      "step": 2286080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.256182179468449e-05,
+      "loss": 3.8592,
+      "step": 2286592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.255343584717397e-05,
+      "loss": 3.8498,
+      "step": 2287104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.254504989966345e-05,
+      "loss": 3.8575,
+      "step": 2287616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.253666395215293e-05,
+      "loss": 3.849,
+      "step": 2288128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2528278004642408e-05,
+      "loss": 3.8628,
+      "step": 2288640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.251990843593562e-05,
+      "loss": 3.8555,
+      "step": 2289152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9764981269836426,
+      "eval_runtime": 290.4494,
+      "eval_samples_per_second": 1313.795,
+      "eval_steps_per_second": 41.057,
+      "step": 2289600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.25115224884251e-05,
+      "loss": 3.8611,
+      "step": 2289664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2503136540914582e-05,
+      "loss": 3.8473,
+      "step": 2290176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.249475059340406e-05,
+      "loss": 3.8516,
+      "step": 2290688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2486381024697271e-05,
+      "loss": 3.8637,
+      "step": 2291200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2477995077186751e-05,
+      "loss": 3.8568,
+      "step": 2291712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2469609129676231e-05,
+      "loss": 3.8565,
+      "step": 2292224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.246122318216571e-05,
+      "loss": 3.8545,
+      "step": 2292736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.245283723465519e-05,
+      "loss": 3.8496,
+      "step": 2293248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.244445128714467e-05,
+      "loss": 3.8542,
+      "step": 2293760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.243606533963415e-05,
+      "loss": 3.8648,
+      "step": 2294272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.242767939212363e-05,
+      "loss": 3.8596,
+      "step": 2294784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2419309823416841e-05,
+      "loss": 3.8581,
+      "step": 2295296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2410940254710054e-05,
+      "loss": 3.8642,
+      "step": 2295808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2402554307199534e-05,
+      "loss": 3.8389,
+      "step": 2296320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2394168359689012e-05,
+      "loss": 3.8522,
+      "step": 2296832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2385782412178492e-05,
+      "loss": 3.8457,
+      "step": 2297344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2377396464667974e-05,
+      "loss": 3.8579,
+      "step": 2297856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2369010517157454e-05,
+      "loss": 3.8546,
+      "step": 2298368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2360624569646932e-05,
+      "loss": 3.8545,
+      "step": 2298880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2352238622136412e-05,
+      "loss": 3.8551,
+      "step": 2299392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2343869053429624e-05,
+      "loss": 3.8747,
+      "step": 2299904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2335483105919103e-05,
+      "loss": 3.8577,
+      "step": 2300416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2327113537212315e-05,
+      "loss": 3.8599,
+      "step": 2300928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2318727589701795e-05,
+      "loss": 3.8615,
+      "step": 2301440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2310341642191275e-05,
+      "loss": 3.8695,
+      "step": 2301952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2301955694680755e-05,
+      "loss": 3.848,
+      "step": 2302464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2293569747170235e-05,
+      "loss": 3.8525,
+      "step": 2302976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2285183799659715e-05,
+      "loss": 3.8618,
+      "step": 2303488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2276797852149193e-05,
+      "loss": 3.8484,
+      "step": 2304000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2268428283442406e-05,
+      "loss": 3.8555,
+      "step": 2304512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2260042335931886e-05,
+      "loss": 3.8631,
+      "step": 2305024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2251656388421366e-05,
+      "loss": 3.8642,
+      "step": 2305536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2243270440910846e-05,
+      "loss": 3.8619,
+      "step": 2306048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2234884493400325e-05,
+      "loss": 3.8656,
+      "step": 2306560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2226498545889805e-05,
+      "loss": 3.8545,
+      "step": 2307072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2218128977183018e-05,
+      "loss": 3.8652,
+      "step": 2307584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2209759408476227e-05,
+      "loss": 3.8581,
+      "step": 2308096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2201373460965709e-05,
+      "loss": 3.8397,
+      "step": 2308608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2192987513455189e-05,
+      "loss": 3.8604,
+      "step": 2309120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2184601565944667e-05,
+      "loss": 3.847,
+      "step": 2309632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2176215618434147e-05,
+      "loss": 3.8467,
+      "step": 2310144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2167829670923628e-05,
+      "loss": 3.8567,
+      "step": 2310656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2159443723413108e-05,
+      "loss": 3.8647,
+      "step": 2311168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2151057775902587e-05,
+      "loss": 3.8473,
+      "step": 2311680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2142671828392067e-05,
+      "loss": 3.863,
+      "step": 2312192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2134285880881548e-05,
+      "loss": 3.8582,
+      "step": 2312704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2125899933371027e-05,
+      "loss": 3.8563,
+      "step": 2313216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2117530364664239e-05,
+      "loss": 3.8599,
+      "step": 2313728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2109144417153719e-05,
+      "loss": 3.8413,
+      "step": 2314240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2100758469643197e-05,
+      "loss": 3.8548,
+      "step": 2314752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2092372522132677e-05,
+      "loss": 3.8567,
+      "step": 2315264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2083986574622159e-05,
+      "loss": 3.8609,
+      "step": 2315776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.207561700591537e-05,
+      "loss": 3.857,
+      "step": 2316288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2067231058404848e-05,
+      "loss": 3.8423,
+      "step": 2316800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.205884511089433e-05,
+      "loss": 3.8486,
+      "step": 2317312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.205045916338381e-05,
+      "loss": 3.8512,
+      "step": 2317824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2042073215873288e-05,
+      "loss": 3.8673,
+      "step": 2318336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2033687268362768e-05,
+      "loss": 3.8368,
+      "step": 2318848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.202531769965598e-05,
+      "loss": 3.8627,
+      "step": 2319360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.201693175214546e-05,
+      "loss": 3.8555,
+      "step": 2319872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.200854580463494e-05,
+      "loss": 3.8383,
+      "step": 2320384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.200015985712442e-05,
+      "loss": 3.8578,
+      "step": 2320896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.19917739096139e-05,
+      "loss": 3.8497,
+      "step": 2321408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1983387962103378e-05,
+      "loss": 3.8391,
+      "step": 2321920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.197500201459286e-05,
+      "loss": 3.8402,
+      "step": 2322432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.196663244588607e-05,
+      "loss": 3.8629,
+      "step": 2322944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1958279255983014e-05,
+      "loss": 3.8531,
+      "step": 2323456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1949893308472494e-05,
+      "loss": 3.8485,
+      "step": 2323968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1941507360961974e-05,
+      "loss": 3.8407,
+      "step": 2324480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1933121413451454e-05,
+      "loss": 3.844,
+      "step": 2324992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1924735465940934e-05,
+      "loss": 3.854,
+      "step": 2325504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1916349518430412e-05,
+      "loss": 3.8593,
+      "step": 2326016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1907963570919894e-05,
+      "loss": 3.856,
+      "step": 2326528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1899577623409374e-05,
+      "loss": 3.8619,
+      "step": 2327040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1891191675898852e-05,
+      "loss": 3.859,
+      "step": 2327552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1882838485995795e-05,
+      "loss": 3.857,
+      "step": 2328064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1874452538485275e-05,
+      "loss": 3.8531,
+      "step": 2328576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1866066590974755e-05,
+      "loss": 3.8481,
+      "step": 2329088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1857680643464235e-05,
+      "loss": 3.8507,
+      "step": 2329600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1849294695953715e-05,
+      "loss": 3.8507,
+      "step": 2330112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1840908748443195e-05,
+      "loss": 3.8524,
+      "step": 2330624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1832522800932675e-05,
+      "loss": 3.8606,
+      "step": 2331136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1824136853422155e-05,
+      "loss": 3.8664,
+      "step": 2331648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1815750905911635e-05,
+      "loss": 3.8545,
+      "step": 2332160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1807381337204847e-05,
+      "loss": 3.8369,
+      "step": 2332672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1798995389694326e-05,
+      "loss": 3.8565,
+      "step": 2333184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1790609442183806e-05,
+      "loss": 3.858,
+      "step": 2333696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1782223494673286e-05,
+      "loss": 3.8582,
+      "step": 2334208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1773837547162767e-05,
+      "loss": 3.852,
+      "step": 2334720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1765467978455976e-05,
+      "loss": 3.8694,
+      "step": 2335232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1757082030945456e-05,
+      "loss": 3.8567,
+      "step": 2335744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1748696083434938e-05,
+      "loss": 3.8502,
+      "step": 2336256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1740326514728149e-05,
+      "loss": 3.852,
+      "step": 2336768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1731940567217629e-05,
+      "loss": 3.8623,
+      "step": 2337280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1723554619707109e-05,
+      "loss": 3.855,
+      "step": 2337792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1715168672196589e-05,
+      "loss": 3.861,
+      "step": 2338304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1706782724686067e-05,
+      "loss": 3.8501,
+      "step": 2338816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1698396777175549e-05,
+      "loss": 3.8458,
+      "step": 2339328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1690010829665029e-05,
+      "loss": 3.8527,
+      "step": 2339840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1681624882154507e-05,
+      "loss": 3.865,
+      "step": 2340352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1673238934643987e-05,
+      "loss": 3.86,
+      "step": 2340864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1664852987133468e-05,
+      "loss": 3.8499,
+      "step": 2341376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1656467039622948e-05,
+      "loss": 3.8522,
+      "step": 2341888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1648081092112427e-05,
+      "loss": 3.8449,
+      "step": 2342400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1639711523405639e-05,
+      "loss": 3.8594,
+      "step": 2342912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.163134195469885e-05,
+      "loss": 3.8479,
+      "step": 2343424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.162295600718833e-05,
+      "loss": 3.8503,
+      "step": 2343936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.161457005967781e-05,
+      "loss": 3.8527,
+      "step": 2344448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.160618411216729e-05,
+      "loss": 3.8549,
+      "step": 2344960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.159779816465677e-05,
+      "loss": 3.8531,
+      "step": 2345472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.158941221714625e-05,
+      "loss": 3.8505,
+      "step": 2345984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.158102626963573e-05,
+      "loss": 3.8475,
+      "step": 2346496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.157264032212521e-05,
+      "loss": 3.8586,
+      "step": 2347008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1564270753418422e-05,
+      "loss": 3.8579,
+      "step": 2347520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.15558848059079e-05,
+      "loss": 3.8549,
+      "step": 2348032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.154749885839738e-05,
+      "loss": 3.8597,
+      "step": 2348544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.153911291088686e-05,
+      "loss": 3.8628,
+      "step": 2349056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1530759720983804e-05,
+      "loss": 3.8551,
+      "step": 2349568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1522373773473284e-05,
+      "loss": 3.8471,
+      "step": 2350080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1513987825962763e-05,
+      "loss": 3.8573,
+      "step": 2350592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1505601878452243e-05,
+      "loss": 3.8501,
+      "step": 2351104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1497215930941722e-05,
+      "loss": 3.8455,
+      "step": 2351616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1488829983431203e-05,
+      "loss": 3.8531,
+      "step": 2352128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1480444035920683e-05,
+      "loss": 3.8495,
+      "step": 2352640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1472058088410162e-05,
+      "loss": 3.857,
+      "step": 2353152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1463688519703374e-05,
+      "loss": 3.8426,
+      "step": 2353664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1455302572192854e-05,
+      "loss": 3.8527,
+      "step": 2354176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1446916624682334e-05,
+      "loss": 3.8473,
+      "step": 2354688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1438547055975545e-05,
+      "loss": 3.8494,
+      "step": 2355200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1430161108465025e-05,
+      "loss": 3.8587,
+      "step": 2355712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1421775160954505e-05,
+      "loss": 3.8612,
+      "step": 2356224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1413389213443985e-05,
+      "loss": 3.8556,
+      "step": 2356736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1405003265933465e-05,
+      "loss": 3.8467,
+      "step": 2357248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1396617318422945e-05,
+      "loss": 3.8619,
+      "step": 2357760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1388231370912424e-05,
+      "loss": 3.8568,
+      "step": 2358272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1379845423401904e-05,
+      "loss": 3.8579,
+      "step": 2358784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1371475854695115e-05,
+      "loss": 3.8604,
+      "step": 2359296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1363089907184595e-05,
+      "loss": 3.8546,
+      "step": 2359808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1354703959674075e-05,
+      "loss": 3.8661,
+      "step": 2360320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1346318012163555e-05,
+      "loss": 3.8623,
+      "step": 2360832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1337932064653035e-05,
+      "loss": 3.8465,
+      "step": 2361344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1329546117142515e-05,
+      "loss": 3.8519,
+      "step": 2361856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1321160169631995e-05,
+      "loss": 3.8529,
+      "step": 2362368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1312774222121475e-05,
+      "loss": 3.853,
+      "step": 2362880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1304404653414687e-05,
+      "loss": 3.8497,
+      "step": 2363392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1296035084707898e-05,
+      "loss": 3.8519,
+      "step": 2363904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1287665516001109e-05,
+      "loss": 3.8461,
+      "step": 2364416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1279279568490589e-05,
+      "loss": 3.8599,
+      "step": 2364928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1270893620980069e-05,
+      "loss": 3.8522,
+      "step": 2365440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.975809335708618,
+      "eval_runtime": 287.921,
+      "eval_samples_per_second": 1325.332,
+      "eval_steps_per_second": 41.418,
+      "step": 2365920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1262507673469549e-05,
+      "loss": 3.8116,
+      "step": 2365952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1254121725959029e-05,
+      "loss": 3.8506,
+      "step": 2366464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1245735778448509e-05,
+      "loss": 3.8442,
+      "step": 2366976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1237349830937989e-05,
+      "loss": 3.8594,
+      "step": 2367488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1228963883427469e-05,
+      "loss": 3.856,
+      "step": 2368000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1220577935916949e-05,
+      "loss": 3.8551,
+      "step": 2368512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1212191988406429e-05,
+      "loss": 3.8494,
+      "step": 2369024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1203806040895907e-05,
+      "loss": 3.844,
+      "step": 2369536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1195420093385388e-05,
+      "loss": 3.8547,
+      "step": 2370048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1187034145874868e-05,
+      "loss": 3.8567,
+      "step": 2370560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1178648198364347e-05,
+      "loss": 3.8565,
+      "step": 2371072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1170262250853827e-05,
+      "loss": 3.8522,
+      "step": 2371584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1161876303343308e-05,
+      "loss": 3.8638,
+      "step": 2372096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1153490355832788e-05,
+      "loss": 3.8355,
+      "step": 2372608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1145104408322267e-05,
+      "loss": 3.8498,
+      "step": 2373120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1136718460811746e-05,
+      "loss": 3.841,
+      "step": 2373632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1128332513301228e-05,
+      "loss": 3.8524,
+      "step": 2374144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1119946565790706e-05,
+      "loss": 3.8529,
+      "step": 2374656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1111560618280186e-05,
+      "loss": 3.8488,
+      "step": 2375168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1103174670769666e-05,
+      "loss": 3.8492,
+      "step": 2375680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1094805102062879e-05,
+      "loss": 3.8715,
+      "step": 2376192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1086419154552357e-05,
+      "loss": 3.8542,
+      "step": 2376704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1078033207041839e-05,
+      "loss": 3.8529,
+      "step": 2377216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1069647259531319e-05,
+      "loss": 3.8614,
+      "step": 2377728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1061277690824528e-05,
+      "loss": 3.865,
+      "step": 2378240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.105290812211774e-05,
+      "loss": 3.847,
+      "step": 2378752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.104452217460722e-05,
+      "loss": 3.8458,
+      "step": 2379264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.10361362270967e-05,
+      "loss": 3.8586,
+      "step": 2379776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.102775027958618e-05,
+      "loss": 3.8437,
+      "step": 2380288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.101936433207566e-05,
+      "loss": 3.8519,
+      "step": 2380800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.101097838456514e-05,
+      "loss": 3.8553,
+      "step": 2381312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.100259243705462e-05,
+      "loss": 3.8614,
+      "step": 2381824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.09942064895441e-05,
+      "loss": 3.8579,
+      "step": 2382336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.098582054203358e-05,
+      "loss": 3.8573,
+      "step": 2382848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0977467352130523e-05,
+      "loss": 3.8514,
+      "step": 2383360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0969097783423734e-05,
+      "loss": 3.8632,
+      "step": 2383872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0960711835913214e-05,
+      "loss": 3.8547,
+      "step": 2384384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0952325888402694e-05,
+      "loss": 3.8376,
+      "step": 2384896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0943939940892174e-05,
+      "loss": 3.856,
+      "step": 2385408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0935553993381654e-05,
+      "loss": 3.8428,
+      "step": 2385920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0927168045871134e-05,
+      "loss": 3.8475,
+      "step": 2386432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0918782098360614e-05,
+      "loss": 3.8514,
+      "step": 2386944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0910396150850092e-05,
+      "loss": 3.8594,
+      "step": 2387456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0902010203339574e-05,
+      "loss": 3.8477,
+      "step": 2387968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0893640634632784e-05,
+      "loss": 3.8536,
+      "step": 2388480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0885254687122264e-05,
+      "loss": 3.8592,
+      "step": 2388992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0876868739611744e-05,
+      "loss": 3.8455,
+      "step": 2389504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0868482792101224e-05,
+      "loss": 3.861,
+      "step": 2390016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0860096844590704e-05,
+      "loss": 3.8399,
+      "step": 2390528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0851710897080184e-05,
+      "loss": 3.8459,
+      "step": 2391040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0843324949569664e-05,
+      "loss": 3.8562,
+      "step": 2391552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0834955380862875e-05,
+      "loss": 3.8544,
+      "step": 2392064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0826569433352355e-05,
+      "loss": 3.8544,
+      "step": 2392576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0818199864645566e-05,
+      "loss": 3.8377,
+      "step": 2393088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0809813917135046e-05,
+      "loss": 3.8507,
+      "step": 2393600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0801427969624527e-05,
+      "loss": 3.8447,
+      "step": 2394112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0793042022114006e-05,
+      "loss": 3.8682,
+      "step": 2394624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0784656074603486e-05,
+      "loss": 3.8313,
+      "step": 2395136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0776270127092965e-05,
+      "loss": 3.8595,
+      "step": 2395648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0767884179582447e-05,
+      "loss": 3.8527,
+      "step": 2396160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0759498232071925e-05,
+      "loss": 3.8346,
+      "step": 2396672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0751128663365136e-05,
+      "loss": 3.8531,
+      "step": 2397184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0742759094658349e-05,
+      "loss": 3.8489,
+      "step": 2397696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0734373147147829e-05,
+      "loss": 3.8328,
+      "step": 2398208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0725987199637309e-05,
+      "loss": 3.8396,
+      "step": 2398720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0717601252126789e-05,
+      "loss": 3.8608,
+      "step": 2399232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.070923168342e-05,
+      "loss": 3.8496,
+      "step": 2399744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.070084573590948e-05,
+      "loss": 3.8459,
+      "step": 2400256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.069245978839896e-05,
+      "loss": 3.8369,
+      "step": 2400768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.068407384088844e-05,
+      "loss": 3.8404,
+      "step": 2401280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0675687893377919e-05,
+      "loss": 3.8526,
+      "step": 2401792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0667301945867399e-05,
+      "loss": 3.8565,
+      "step": 2402304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0658915998356879e-05,
+      "loss": 3.8506,
+      "step": 2402816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0650530050846359e-05,
+      "loss": 3.8618,
+      "step": 2403328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.064216048213957e-05,
+      "loss": 3.8574,
+      "step": 2403840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.063377453462905e-05,
+      "loss": 3.8489,
+      "step": 2404352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0625404965922262e-05,
+      "loss": 3.8485,
+      "step": 2404864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0617019018411742e-05,
+      "loss": 3.8461,
+      "step": 2405376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.060863307090122e-05,
+      "loss": 3.848,
+      "step": 2405888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.06002471233907e-05,
+      "loss": 3.8512,
+      "step": 2406400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0591861175880182e-05,
+      "loss": 3.8438,
+      "step": 2406912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.058347522836966e-05,
+      "loss": 3.8576,
+      "step": 2407424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.057508928085914e-05,
+      "loss": 3.8561,
+      "step": 2407936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.056670333334862e-05,
+      "loss": 3.8536,
+      "step": 2408448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0558317385838102e-05,
+      "loss": 3.8363,
+      "step": 2408960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0549947817131311e-05,
+      "loss": 3.8481,
+      "step": 2409472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0541561869620793e-05,
+      "loss": 3.8596,
+      "step": 2409984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0533175922110273e-05,
+      "loss": 3.8523,
+      "step": 2410496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0524789974599751e-05,
+      "loss": 3.8502,
+      "step": 2411008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.051640402708923e-05,
+      "loss": 3.8625,
+      "step": 2411520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0508034458382443e-05,
+      "loss": 3.8521,
+      "step": 2412032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0499648510871923e-05,
+      "loss": 3.848,
+      "step": 2412544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0491262563361402e-05,
+      "loss": 3.8502,
+      "step": 2413056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0482876615850883e-05,
+      "loss": 3.8577,
+      "step": 2413568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0474490668340363e-05,
+      "loss": 3.8499,
+      "step": 2414080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0466104720829841e-05,
+      "loss": 3.8646,
+      "step": 2414592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0457718773319321e-05,
+      "loss": 3.8411,
+      "step": 2415104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0449365583416265e-05,
+      "loss": 3.8496,
+      "step": 2415616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0440979635905746e-05,
+      "loss": 3.8486,
+      "step": 2416128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0432610067198957e-05,
+      "loss": 3.8569,
+      "step": 2416640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0424224119688435e-05,
+      "loss": 3.8616,
+      "step": 2417152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0415838172177917e-05,
+      "loss": 3.8413,
+      "step": 2417664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0407452224667397e-05,
+      "loss": 3.8521,
+      "step": 2418176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0399066277156875e-05,
+      "loss": 3.8421,
+      "step": 2418688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0390680329646355e-05,
+      "loss": 3.8534,
+      "step": 2419200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0382294382135837e-05,
+      "loss": 3.8375,
+      "step": 2419712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0373908434625315e-05,
+      "loss": 3.8495,
+      "step": 2420224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0365522487114795e-05,
+      "loss": 3.8502,
+      "step": 2420736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0357152918408008e-05,
+      "loss": 3.8473,
+      "step": 2421248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0348766970897487e-05,
+      "loss": 3.8546,
+      "step": 2421760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0340381023386966e-05,
+      "loss": 3.8497,
+      "step": 2422272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0331995075876447e-05,
+      "loss": 3.8391,
+      "step": 2422784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0323609128365927e-05,
+      "loss": 3.856,
+      "step": 2423296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0315239559659136e-05,
+      "loss": 3.8526,
+      "step": 2423808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0306853612148618e-05,
+      "loss": 3.8513,
+      "step": 2424320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0298467664638098e-05,
+      "loss": 3.857,
+      "step": 2424832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0290081717127578e-05,
+      "loss": 3.8586,
+      "step": 2425344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0281695769617056e-05,
+      "loss": 3.852,
+      "step": 2425856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0273309822106538e-05,
+      "loss": 3.8427,
+      "step": 2426368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0264923874596018e-05,
+      "loss": 3.8536,
+      "step": 2426880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0256537927085496e-05,
+      "loss": 3.8455,
+      "step": 2427392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0248168358378709e-05,
+      "loss": 3.8446,
+      "step": 2427904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0239782410868189e-05,
+      "loss": 3.8478,
+      "step": 2428416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0231396463357669e-05,
+      "loss": 3.8489,
+      "step": 2428928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0223010515847148e-05,
+      "loss": 3.8534,
+      "step": 2429440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0214624568336628e-05,
+      "loss": 3.8416,
+      "step": 2429952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.020625499962984e-05,
+      "loss": 3.8457,
+      "step": 2430464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.019786905211932e-05,
+      "loss": 3.8432,
+      "step": 2430976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.018949948341253e-05,
+      "loss": 3.85,
+      "step": 2431488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.018111353590201e-05,
+      "loss": 3.8506,
+      "step": 2432000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0172727588391492e-05,
+      "loss": 3.8589,
+      "step": 2432512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.016434164088097e-05,
+      "loss": 3.8536,
+      "step": 2433024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.015595569337045e-05,
+      "loss": 3.8456,
+      "step": 2433536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0147586124663662e-05,
+      "loss": 3.8537,
+      "step": 2434048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0139200177153142e-05,
+      "loss": 3.8553,
+      "step": 2434560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.013081422964262e-05,
+      "loss": 3.8564,
+      "step": 2435072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0122428282132102e-05,
+      "loss": 3.8616,
+      "step": 2435584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0114042334621582e-05,
+      "loss": 3.8502,
+      "step": 2436096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.010565638711106e-05,
+      "loss": 3.8639,
+      "step": 2436608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.009727043960054e-05,
+      "loss": 3.8552,
+      "step": 2437120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0088884492090022e-05,
+      "loss": 3.8461,
+      "step": 2437632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0080514923383233e-05,
+      "loss": 3.8516,
+      "step": 2438144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0072128975872713e-05,
+      "loss": 3.847,
+      "step": 2438656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0063743028362193e-05,
+      "loss": 3.8487,
+      "step": 2439168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0055373459655403e-05,
+      "loss": 3.8429,
+      "step": 2439680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0046987512144883e-05,
+      "loss": 3.8539,
+      "step": 2440192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0038601564634363e-05,
+      "loss": 3.8421,
+      "step": 2440704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0030215617123843e-05,
+      "loss": 3.8553,
+      "step": 2441216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0021846048417056e-05,
+      "loss": 3.8501,
+      "step": 2441728
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0013460100906534e-05,
+      "loss": 3.8512,
+      "step": 2442240
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9745898246765137,
+      "eval_runtime": 289.2682,
+      "eval_samples_per_second": 1319.16,
+      "eval_steps_per_second": 41.225,
+      "step": 2442240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0005074153396014e-05,
+      "loss": 3.8443,
+      "step": 2442752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.996688205885494e-06,
+      "loss": 3.8432,
+      "step": 2443264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.988318637178706e-06,
+      "loss": 3.854,
+      "step": 2443776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.979932689668185e-06,
+      "loss": 3.8547,
+      "step": 2444288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.971546742157666e-06,
+      "loss": 3.853,
+      "step": 2444800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.963160794647146e-06,
+      "loss": 3.8445,
+      "step": 2445312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.954774847136625e-06,
+      "loss": 3.8425,
+      "step": 2445824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.946388899626105e-06,
+      "loss": 3.845,
+      "step": 2446336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.938002952115586e-06,
+      "loss": 3.8558,
+      "step": 2446848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.929617004605064e-06,
+      "loss": 3.8491,
+      "step": 2447360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.921231057094544e-06,
+      "loss": 3.8555,
+      "step": 2447872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.912845109584024e-06,
+      "loss": 3.8613,
+      "step": 2448384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.904459162073506e-06,
+      "loss": 3.8352,
+      "step": 2448896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.896089593366715e-06,
+      "loss": 3.8409,
+      "step": 2449408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.887703645856195e-06,
+      "loss": 3.8381,
+      "step": 2449920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.879317698345677e-06,
+      "loss": 3.8479,
+      "step": 2450432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.870931750835155e-06,
+      "loss": 3.8509,
+      "step": 2450944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.862545803324635e-06,
+      "loss": 3.849,
+      "step": 2451456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.854159855814115e-06,
+      "loss": 3.8455,
+      "step": 2451968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.845773908303597e-06,
+      "loss": 3.8679,
+      "step": 2452480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.837404339596806e-06,
+      "loss": 3.8514,
+      "step": 2452992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.829018392086287e-06,
+      "loss": 3.8539,
+      "step": 2453504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.820632444575767e-06,
+      "loss": 3.8529,
+      "step": 2454016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.812246497065246e-06,
+      "loss": 3.8654,
+      "step": 2454528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.803860549554725e-06,
+      "loss": 3.8446,
+      "step": 2455040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.795474602044207e-06,
+      "loss": 3.8466,
+      "step": 2455552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.787088654533687e-06,
+      "loss": 3.8531,
+      "step": 2456064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.778719085826898e-06,
+      "loss": 3.8421,
+      "step": 2456576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.770333138316378e-06,
+      "loss": 3.8475,
+      "step": 2457088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.761947190805858e-06,
+      "loss": 3.8512,
+      "step": 2457600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.753577622099069e-06,
+      "loss": 3.8608,
+      "step": 2458112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.745191674588549e-06,
+      "loss": 3.8526,
+      "step": 2458624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.736805727078028e-06,
+      "loss": 3.8546,
+      "step": 2459136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.728419779567508e-06,
+      "loss": 3.853,
+      "step": 2459648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.720033832056988e-06,
+      "loss": 3.856,
+      "step": 2460160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.711647884546468e-06,
+      "loss": 3.8518,
+      "step": 2460672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.703261937035948e-06,
+      "loss": 3.8372,
+      "step": 2461184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.694875989525427e-06,
+      "loss": 3.8487,
+      "step": 2461696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.686490042014908e-06,
+      "loss": 3.8402,
+      "step": 2462208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.678104094504388e-06,
+      "loss": 3.8479,
+      "step": 2462720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.669718146993866e-06,
+      "loss": 3.8469,
+      "step": 2463232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.661348578287079e-06,
+      "loss": 3.8566,
+      "step": 2463744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.652962630776559e-06,
+      "loss": 3.8442,
+      "step": 2464256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.644576683266039e-06,
+      "loss": 3.849,
+      "step": 2464768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.636190735755519e-06,
+      "loss": 3.8571,
+      "step": 2465280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.627804788244999e-06,
+      "loss": 3.8444,
+      "step": 2465792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.619418840734479e-06,
+      "loss": 3.8574,
+      "step": 2466304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.611049272027691e-06,
+      "loss": 3.8343,
+      "step": 2466816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.60266332451717e-06,
+      "loss": 3.8467,
+      "step": 2467328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.59429375581038e-06,
+      "loss": 3.857,
+      "step": 2467840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.585907808299862e-06,
+      "loss": 3.8517,
+      "step": 2468352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.57752186078934e-06,
+      "loss": 3.8494,
+      "step": 2468864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.56913591327882e-06,
+      "loss": 3.8303,
+      "step": 2469376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.5607499657683e-06,
+      "loss": 3.852,
+      "step": 2469888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.552364018257782e-06,
+      "loss": 3.8374,
+      "step": 2470400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.54397807074726e-06,
+      "loss": 3.8636,
+      "step": 2470912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.53559212323674e-06,
+      "loss": 3.8347,
+      "step": 2471424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.52720617572622e-06,
+      "loss": 3.8549,
+      "step": 2471936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.5188202282157e-06,
+      "loss": 3.8444,
+      "step": 2472448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.51043428070518e-06,
+      "loss": 3.8339,
+      "step": 2472960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.50204833319466e-06,
+      "loss": 3.8508,
+      "step": 2473472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.493678764487872e-06,
+      "loss": 3.8434,
+      "step": 2473984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.48529281697735e-06,
+      "loss": 3.8268,
+      "step": 2474496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.47690686946683e-06,
+      "loss": 3.8356,
+      "step": 2475008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.468520921956312e-06,
+      "loss": 3.859,
+      "step": 2475520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.460151353249521e-06,
+      "loss": 3.8439,
+      "step": 2476032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.451781784542734e-06,
+      "loss": 3.8393,
+      "step": 2476544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.443395837032214e-06,
+      "loss": 3.8382,
+      "step": 2477056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.435009889521694e-06,
+      "loss": 3.8352,
+      "step": 2477568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.426623942011174e-06,
+      "loss": 3.8463,
+      "step": 2478080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.418237994500653e-06,
+      "loss": 3.8532,
+      "step": 2478592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.409852046990133e-06,
+      "loss": 3.8484,
+      "step": 2479104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.401482478283346e-06,
+      "loss": 3.8583,
+      "step": 2479616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.393096530772824e-06,
+      "loss": 3.8505,
+      "step": 2480128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.384710583262304e-06,
+      "loss": 3.8517,
+      "step": 2480640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.376324635751784e-06,
+      "loss": 3.8409,
+      "step": 2481152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.367938688241264e-06,
+      "loss": 3.848,
+      "step": 2481664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.359552740730744e-06,
+      "loss": 3.8491,
+      "step": 2482176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.351166793220224e-06,
+      "loss": 3.8418,
+      "step": 2482688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.342780845709704e-06,
+      "loss": 3.8413,
+      "step": 2483200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.334411277002915e-06,
+      "loss": 3.8533,
+      "step": 2483712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.326025329492395e-06,
+      "loss": 3.8574,
+      "step": 2484224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.317639381981876e-06,
+      "loss": 3.848,
+      "step": 2484736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.309253434471355e-06,
+      "loss": 3.8324,
+      "step": 2485248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.300883865764565e-06,
+      "loss": 3.8486,
+      "step": 2485760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.292497918254047e-06,
+      "loss": 3.8524,
+      "step": 2486272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.284111970743527e-06,
+      "loss": 3.8541,
+      "step": 2486784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.275726023233005e-06,
+      "loss": 3.8452,
+      "step": 2487296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.267340075722485e-06,
+      "loss": 3.8614,
+      "step": 2487808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.258954128211967e-06,
+      "loss": 3.85,
+      "step": 2488320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.250568180701445e-06,
+      "loss": 3.841,
+      "step": 2488832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.242182233190925e-06,
+      "loss": 3.8487,
+      "step": 2489344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.233812664484138e-06,
+      "loss": 3.8536,
+      "step": 2489856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.225426716973618e-06,
+      "loss": 3.8458,
+      "step": 2490368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.217040769463096e-06,
+      "loss": 3.8604,
+      "step": 2490880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.208654821952577e-06,
+      "loss": 3.8426,
+      "step": 2491392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.200285253245788e-06,
+      "loss": 3.8405,
+      "step": 2491904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.191899305735266e-06,
+      "loss": 3.844,
+      "step": 2492416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.183513358224748e-06,
+      "loss": 3.8574,
+      "step": 2492928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.175127410714228e-06,
+      "loss": 3.8538,
+      "step": 2493440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.166757842007439e-06,
+      "loss": 3.8406,
+      "step": 2493952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.15838827330065e-06,
+      "loss": 3.8466,
+      "step": 2494464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.15000232579013e-06,
+      "loss": 3.8409,
+      "step": 2494976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.141616378279611e-06,
+      "loss": 3.855,
+      "step": 2495488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.13323043076909e-06,
+      "loss": 3.8313,
+      "step": 2496000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.12484448325857e-06,
+      "loss": 3.8472,
+      "step": 2496512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.11645853574805e-06,
+      "loss": 3.852,
+      "step": 2497024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.108088967041262e-06,
+      "loss": 3.8451,
+      "step": 2497536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.09970301953074e-06,
+      "loss": 3.8481,
+      "step": 2498048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.09131707202022e-06,
+      "loss": 3.8489,
+      "step": 2498560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.082931124509702e-06,
+      "loss": 3.8383,
+      "step": 2499072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.07454517699918e-06,
+      "loss": 3.8496,
+      "step": 2499584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.066175608292393e-06,
+      "loss": 3.8501,
+      "step": 2500096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.057789660781872e-06,
+      "loss": 3.8483,
+      "step": 2500608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.049403713271352e-06,
+      "loss": 3.8534,
+      "step": 2501120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.04101776576083e-06,
+      "loss": 3.8553,
+      "step": 2501632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.032631818250312e-06,
+      "loss": 3.8503,
+      "step": 2502144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.024245870739792e-06,
+      "loss": 3.8432,
+      "step": 2502656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.015876302033003e-06,
+      "loss": 3.8469,
+      "step": 2503168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.007490354522483e-06,
+      "loss": 3.8453,
+      "step": 2503680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.999104407011963e-06,
+      "loss": 3.839,
+      "step": 2504192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.990718459501443e-06,
+      "loss": 3.8433,
+      "step": 2504704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.982332511990923e-06,
+      "loss": 3.8441,
+      "step": 2505216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.973946564480403e-06,
+      "loss": 3.8537,
+      "step": 2505728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.965560616969883e-06,
+      "loss": 3.8389,
+      "step": 2506240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.957174669459361e-06,
+      "loss": 3.8441,
+      "step": 2506752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.948805100752574e-06,
+      "loss": 3.8409,
+      "step": 2507264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.940419153242054e-06,
+      "loss": 3.8471,
+      "step": 2507776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.932049584535266e-06,
+      "loss": 3.8396,
+      "step": 2508288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.923663637024744e-06,
+      "loss": 3.861,
+      "step": 2508800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.915277689514224e-06,
+      "loss": 3.8494,
+      "step": 2509312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.906891742003704e-06,
+      "loss": 3.8445,
+      "step": 2509824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.898522173296917e-06,
+      "loss": 3.8493,
+      "step": 2510336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.890136225786395e-06,
+      "loss": 3.8478,
+      "step": 2510848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.881750278275875e-06,
+      "loss": 3.8548,
+      "step": 2511360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.873364330765357e-06,
+      "loss": 3.8565,
+      "step": 2511872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.864978383254835e-06,
+      "loss": 3.8497,
+      "step": 2512384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.856592435744315e-06,
+      "loss": 3.8615,
+      "step": 2512896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.848206488233796e-06,
+      "loss": 3.851,
+      "step": 2513408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.839820540723276e-06,
+      "loss": 3.8447,
+      "step": 2513920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.831450972016485e-06,
+      "loss": 3.8432,
+      "step": 2514432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.823065024505967e-06,
+      "loss": 3.8486,
+      "step": 2514944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.814679076995447e-06,
+      "loss": 3.8464,
+      "step": 2515456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.806293129484925e-06,
+      "loss": 3.838,
+      "step": 2515968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.797923560778138e-06,
+      "loss": 3.8476,
+      "step": 2516480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.789537613267618e-06,
+      "loss": 3.843,
+      "step": 2516992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.781151665757098e-06,
+      "loss": 3.8527,
+      "step": 2517504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.772782097050309e-06,
+      "loss": 3.8477,
+      "step": 2518016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.764396149539788e-06,
+      "loss": 3.8452,
+      "step": 2518528
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.973715305328369,
+      "eval_runtime": 288.7725,
+      "eval_samples_per_second": 1321.424,
+      "eval_steps_per_second": 41.295,
+      "step": 2518560
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.0338091713711717e+18,
+  "trial_name": null,
+  "trial_params": null
+}