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": 3.979726791381836,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/passive/lstm/1/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": 10.8184,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5501,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0545,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9925,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.945,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8588,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7022,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6044,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5141,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4386,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3794,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3101,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2455,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1837,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1337,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0701,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0368,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9885,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.948,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9079,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8763,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8376,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8103,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7674,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7556,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7158,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6974,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.657,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.641,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6204,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.5951,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5853,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5601,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5426,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5344,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5123,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5029,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.458,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 5.463,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.4249,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4319,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4097,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4072,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.377,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.384,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3734,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3663,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3577,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3286,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3204,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3207,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3066,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.301,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.278,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.274,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.255,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2773,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2302,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2461,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2351,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2166,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2139,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948020228477761e-05,
+      "loss": 5.2009,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.1898,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.1832,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.1968,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.1724,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1618,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1519,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1495,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1569,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1505,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.126,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1407,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1351,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1067,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1056,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.0965,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.0877,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0835,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0878,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0828,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0852,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0608,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.043,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0494,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.0504,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.049,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0319,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0311,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0306,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0214,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0126,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0092,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 4.9969,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0083,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.9975,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9754,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9812,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.9784,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.9857,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9668,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9652,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9499,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9566,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9345,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.9463,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9332,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9359,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.9245,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.9296,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906945464479944e-05,
+      "loss": 4.9151,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9195,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9108,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9047,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9159,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.9011,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019155338540045e-05,
+      "loss": 4.8993,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010769391029525e-05,
+      "loss": 4.8892,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002383443519005e-05,
+      "loss": 4.8935,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.8894,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89856279273017e-05,
+      "loss": 4.8819,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897724197979118e-05,
+      "loss": 4.8858,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896885603228066e-05,
+      "loss": 4.8724,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8733,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8538,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8582,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8607,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8507,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8436,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8651,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.847,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8399,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8397,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.834,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8367,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8303,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8271,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8299,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8286,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8168,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.8068,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.8182,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8051,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8073,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.7982,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.7976,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.8037,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.7901,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87509032910258e-05,
+      "loss": 4.786,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.754786968231201,
+      "eval_runtime": 288.6597,
+      "eval_samples_per_second": 1321.941,
+      "eval_steps_per_second": 41.312,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.7731,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.7718,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.7983,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871737587978745e-05,
+      "loss": 4.7815,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870898993227693e-05,
+      "loss": 4.7726,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87006039847664e-05,
+      "loss": 4.7693,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869221803725588e-05,
+      "loss": 4.7701,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868383208974536e-05,
+      "loss": 4.7527,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867544614223484e-05,
+      "loss": 4.7624,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.7612,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658690626017536e-05,
+      "loss": 4.7634,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650304678507016e-05,
+      "loss": 4.762,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8641935109800225e-05,
+      "loss": 4.7437,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633549162289705e-05,
+      "loss": 4.7441,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625163214779185e-05,
+      "loss": 4.733,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616777267268665e-05,
+      "loss": 4.7355,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608391319758145e-05,
+      "loss": 4.7408,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600005372247625e-05,
+      "loss": 4.7314,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591619424737105e-05,
+      "loss": 4.7363,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583233477226585e-05,
+      "loss": 4.7499,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574847529716065e-05,
+      "loss": 4.7328,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566477961009274e-05,
+      "loss": 4.7344,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558092013498754e-05,
+      "loss": 4.7191,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854970606598824e-05,
+      "loss": 4.7366,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854132011847772e-05,
+      "loss": 4.7156,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.718,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852456460226041e-05,
+      "loss": 4.705,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851617865474989e-05,
+      "loss": 4.7135,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85078090860431e-05,
+      "loss": 4.7049,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849942313853258e-05,
+      "loss": 4.7011,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849103719102206e-05,
+      "loss": 4.7059,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848265124351154e-05,
+      "loss": 4.7102,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847426529600102e-05,
+      "loss": 4.7009,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84658793484905e-05,
+      "loss": 4.703,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845749340097998e-05,
+      "loss": 4.6937,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844910745346946e-05,
+      "loss": 4.7033,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440737884762674e-05,
+      "loss": 4.675,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432351937252154e-05,
+      "loss": 4.6904,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423965989741634e-05,
+      "loss": 4.6668,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415580042231114e-05,
+      "loss": 4.6871,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840719409472059e-05,
+      "loss": 4.6712,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839880814721007e-05,
+      "loss": 4.6835,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839042219969955e-05,
+      "loss": 4.6615,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838203625218903e-05,
+      "loss": 4.6803,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837365030467851e-05,
+      "loss": 4.6777,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836526435716799e-05,
+      "loss": 4.675,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835687840965747e-05,
+      "loss": 4.6805,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834849246214695e-05,
+      "loss": 4.6515,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834010651463643e-05,
+      "loss": 4.6524,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833172056712591e-05,
+      "loss": 4.6678,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832333461961539e-05,
+      "loss": 4.6598,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83149650509086e-05,
+      "loss": 4.6566,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830657910339808e-05,
+      "loss": 4.6501,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829820953469129e-05,
+      "loss": 4.6443,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828982358718077e-05,
+      "loss": 4.6382,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828145401847398e-05,
+      "loss": 4.6546,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827306807096346e-05,
+      "loss": 4.6313,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826468212345294e-05,
+      "loss": 4.6399,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825629617594242e-05,
+      "loss": 4.6469,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82479102284319e-05,
+      "loss": 4.6303,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823952428092138e-05,
+      "loss": 4.6303,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823113833341086e-05,
+      "loss": 4.6279,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822275238590034e-05,
+      "loss": 4.6177,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214382817193557e-05,
+      "loss": 4.6221,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8205996869683037e-05,
+      "loss": 4.6382,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197610922172516e-05,
+      "loss": 4.6211,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189224974661996e-05,
+      "loss": 4.6147,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180839027151476e-05,
+      "loss": 4.614,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817245307964095e-05,
+      "loss": 4.6145,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816406713213043e-05,
+      "loss": 4.6294,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155697563423645e-05,
+      "loss": 4.6211,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147311615913125e-05,
+      "loss": 4.611,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81389256684026e-05,
+      "loss": 4.6251,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813053972089208e-05,
+      "loss": 4.6229,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122153773381565e-05,
+      "loss": 4.6072,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113767825871045e-05,
+      "loss": 4.6047,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105381878360525e-05,
+      "loss": 4.604,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8096995930850005e-05,
+      "loss": 4.6005,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8088626362143214e-05,
+      "loss": 4.5957,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8080240414632694e-05,
+      "loss": 4.5998,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071854467122174e-05,
+      "loss": 4.6067,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063468519611654e-05,
+      "loss": 4.6087,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055082572101134e-05,
+      "loss": 4.5966,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046696624590614e-05,
+      "loss": 4.5738,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803832705588382e-05,
+      "loss": 4.5896,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80299411083733e-05,
+      "loss": 4.5936,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802155516086278e-05,
+      "loss": 4.5955,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801316921335226e-05,
+      "loss": 4.5868,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800479964464548e-05,
+      "loss": 4.5843,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799641369713496e-05,
+      "loss": 4.5899,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798802774962444e-05,
+      "loss": 4.5813,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797964180211392e-05,
+      "loss": 4.5765,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797127223340713e-05,
+      "loss": 4.583,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796290266470034e-05,
+      "loss": 4.5686,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795451671718982e-05,
+      "loss": 4.5871,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79461307696793e-05,
+      "loss": 4.5782,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793774482216878e-05,
+      "loss": 4.5585,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792935887465826e-05,
+      "loss": 4.5701,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792098930595147e-05,
+      "loss": 4.569,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791260335844095e-05,
+      "loss": 4.5797,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790421741093043e-05,
+      "loss": 4.5673,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789583146341991e-05,
+      "loss": 4.5631,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788744551590939e-05,
+      "loss": 4.5549,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787905956839887e-05,
+      "loss": 4.5665,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787067362088835e-05,
+      "loss": 4.548,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786230405218156e-05,
+      "loss": 4.5558,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785391810467104e-05,
+      "loss": 4.5531,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784553215716052e-05,
+      "loss": 4.5542,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783714620965e-05,
+      "loss": 4.5447,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782876026213948e-05,
+      "loss": 4.5536,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782037431462896e-05,
+      "loss": 4.5455,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781200474592217e-05,
+      "loss": 4.5458,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780361879841165e-05,
+      "loss": 4.5487,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779523285090114e-05,
+      "loss": 4.5381,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778684690339061e-05,
+      "loss": 4.5632,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777846095588009e-05,
+      "loss": 4.5478,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777007500836957e-05,
+      "loss": 4.5451,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776168906085905e-05,
+      "loss": 4.5365,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775330311334853e-05,
+      "loss": 4.543,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774491716583801e-05,
+      "loss": 4.542,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773653121832749e-05,
+      "loss": 4.5318,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772814527081697e-05,
+      "loss": 4.5475,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771975932330645e-05,
+      "loss": 4.5358,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771138975459966e-05,
+      "loss": 4.5307,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703020185892875e-05,
+      "loss": 4.5256,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694634238382355e-05,
+      "loss": 4.5272,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686248290871835e-05,
+      "loss": 4.525,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7677862343361315e-05,
+      "loss": 4.5277,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7669476395850795e-05,
+      "loss": 4.5143,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7661090448340274e-05,
+      "loss": 4.5431,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7652704500829754e-05,
+      "loss": 4.5322,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7644318553319234e-05,
+      "loss": 4.5226,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7635948984612444e-05,
+      "loss": 4.5202,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7627563037101923e-05,
+      "loss": 4.5157,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7619177089591403e-05,
+      "loss": 4.529,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761079114208088e-05,
+      "loss": 4.5234,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760240519457036e-05,
+      "loss": 4.5189,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759401924705984e-05,
+      "loss": 4.5249,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758563329954932e-05,
+      "loss": 4.5213,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75772473520388e-05,
+      "loss": 4.5196,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756887778333202e-05,
+      "loss": 4.5094,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75604918358215e-05,
+      "loss": 4.5231,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755210588831097e-05,
+      "loss": 4.5106,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754373631960419e-05,
+      "loss": 4.5135,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753535037209367e-05,
+      "loss": 4.5085,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752696442458315e-05,
+      "loss": 4.5097,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751857847707262e-05,
+      "loss": 4.51,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75101925295621e-05,
+      "loss": 4.4987,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501839339659046e-05,
+      "loss": 4.5038,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.476966381072998,
+      "eval_runtime": 289.007,
+      "eval_samples_per_second": 1320.352,
+      "eval_steps_per_second": 41.262,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493453392148526e-05,
+      "loss": 4.4916,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748506744463801e-05,
+      "loss": 4.4895,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747668149712749e-05,
+      "loss": 4.5162,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746829554961697e-05,
+      "loss": 4.5077,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459909602106446e-05,
+      "loss": 4.4957,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451523654595926e-05,
+      "loss": 4.494,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443137707085406e-05,
+      "loss": 4.4976,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434751759574886e-05,
+      "loss": 4.4819,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426365812064366e-05,
+      "loss": 4.4996,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417979864553846e-05,
+      "loss": 4.4941,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409593917043326e-05,
+      "loss": 4.4959,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401207969532806e-05,
+      "loss": 4.4974,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392838400826015e-05,
+      "loss": 4.4802,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384452453315495e-05,
+      "loss": 4.4842,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376066505804975e-05,
+      "loss": 4.478,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736768055829446e-05,
+      "loss": 4.4789,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735929461078394e-05,
+      "loss": 4.483,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735090866327342e-05,
+      "loss": 4.4751,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734253909456663e-05,
+      "loss": 4.482,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733415314705611e-05,
+      "loss": 4.5014,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732576719954559e-05,
+      "loss": 4.4807,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731738125203507e-05,
+      "loss": 4.4834,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730899530452455e-05,
+      "loss": 4.476,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730060935701403e-05,
+      "loss": 4.4876,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729222340950351e-05,
+      "loss": 4.4692,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728383746199298e-05,
+      "loss": 4.4792,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72754678932862e-05,
+      "loss": 4.4634,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726708194577568e-05,
+      "loss": 4.4682,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725869599826516e-05,
+      "loss": 4.466,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725031005075464e-05,
+      "loss": 4.4621,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724192410324412e-05,
+      "loss": 4.468,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233554534537335e-05,
+      "loss": 4.4761,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722516858702681e-05,
+      "loss": 4.4672,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721678263951629e-05,
+      "loss": 4.4731,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720839669200577e-05,
+      "loss": 4.4577,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720001074449525e-05,
+      "loss": 4.4756,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719164117578846e-05,
+      "loss": 4.4433,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718325522827794e-05,
+      "loss": 4.4666,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717486928076742e-05,
+      "loss": 4.4403,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71664833332569e-05,
+      "loss": 4.4619,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158097385746384e-05,
+      "loss": 4.4489,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714972781703959e-05,
+      "loss": 4.4628,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714134186952907e-05,
+      "loss": 4.4434,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713295592201855e-05,
+      "loss": 4.4561,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712456997450803e-05,
+      "loss": 4.4566,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711618402699751e-05,
+      "loss": 4.4581,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710779807948699e-05,
+      "loss": 4.4588,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709941213197647e-05,
+      "loss": 4.4383,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709102618446595e-05,
+      "loss": 4.4343,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708265661575916e-05,
+      "loss": 4.4567,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707427066824864e-05,
+      "loss": 4.4489,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706588472073812e-05,
+      "loss": 4.4423,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70574987732276e-05,
+      "loss": 4.4402,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704912920452082e-05,
+      "loss": 4.4366,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040759635814026e-05,
+      "loss": 4.4269,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032390067107235e-05,
+      "loss": 4.4436,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024004119596715e-05,
+      "loss": 4.4251,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.435,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4452,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4265,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.4284,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982074382044115e-05,
+      "loss": 4.431,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973688434533595e-05,
+      "loss": 4.4159,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965302487023075e-05,
+      "loss": 4.417,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956916539512555e-05,
+      "loss": 4.4394,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948530592002035e-05,
+      "loss": 4.4241,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694016102329525e-05,
+      "loss": 4.4178,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693177507578473e-05,
+      "loss": 4.4176,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692338912827421e-05,
+      "loss": 4.4227,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691500318076369e-05,
+      "loss": 4.4338,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690661723325317e-05,
+      "loss": 4.4301,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898231285742644e-05,
+      "loss": 4.4183,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6889845338232124e-05,
+      "loss": 4.4327,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881459390721604e-05,
+      "loss": 4.4284,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687308982201481e-05,
+      "loss": 4.4168,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686470387450429e-05,
+      "loss": 4.4143,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685631792699377e-05,
+      "loss": 4.4209,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684794835828699e-05,
+      "loss": 4.414,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683956241077647e-05,
+      "loss": 4.4137,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683117646326595e-05,
+      "loss": 4.413,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682279051575543e-05,
+      "loss": 4.4203,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681440456824491e-05,
+      "loss": 4.4261,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680601862073439e-05,
+      "loss": 4.4115,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679763267322387e-05,
+      "loss": 4.3964,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678926310451708e-05,
+      "loss": 4.4063,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678087715700656e-05,
+      "loss": 4.4083,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677249120949604e-05,
+      "loss": 4.416,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676410526198552e-05,
+      "loss": 4.4068,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755735693278726e-05,
+      "loss": 4.4057,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747349745768206e-05,
+      "loss": 4.416,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673896379825769e-05,
+      "loss": 4.4005,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673057785074717e-05,
+      "loss": 4.4007,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672220828204038e-05,
+      "loss": 4.4055,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671382233452986e-05,
+      "loss": 4.3978,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670543638701934e-05,
+      "loss": 4.4146,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669705043950882e-05,
+      "loss": 4.4001,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66886644919983e-05,
+      "loss": 4.3875,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668027854448778e-05,
+      "loss": 4.3986,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667189259697726e-05,
+      "loss": 4.3965,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666350664946674e-05,
+      "loss": 4.4123,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665513708075995e-05,
+      "loss": 4.4021,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664675113324943e-05,
+      "loss": 4.3926,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663836518573891e-05,
+      "loss": 4.3844,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662997923822839e-05,
+      "loss": 4.3963,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662160966952161e-05,
+      "loss": 4.3812,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661322372201109e-05,
+      "loss": 4.3917,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604837774500567e-05,
+      "loss": 4.3866,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596451826990047e-05,
+      "loss": 4.3894,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588065879479526e-05,
+      "loss": 4.3777,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579679931969006e-05,
+      "loss": 4.3965,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657129398445848e-05,
+      "loss": 4.3806,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562924415751695e-05,
+      "loss": 4.3812,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554538468241175e-05,
+      "loss": 4.389,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654615252073065e-05,
+      "loss": 4.3771,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653776657322013e-05,
+      "loss": 4.4001,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529380625709615e-05,
+      "loss": 4.3877,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652101105700283e-05,
+      "loss": 4.3878,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512625109492304e-05,
+      "loss": 4.3784,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504239161981784e-05,
+      "loss": 4.3837,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495853214471264e-05,
+      "loss": 4.3846,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487467266960744e-05,
+      "loss": 4.3729,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479081319450224e-05,
+      "loss": 4.3886,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647071175074343e-05,
+      "loss": 4.3812,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646232580323291e-05,
+      "loss": 4.3755,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645393985572239e-05,
+      "loss": 4.3751,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644555390821187e-05,
+      "loss": 4.3748,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643716796070135e-05,
+      "loss": 4.3652,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642879839199457e-05,
+      "loss": 4.379,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642041244448405e-05,
+      "loss": 4.3636,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641202649697353e-05,
+      "loss": 4.3863,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640364054946301e-05,
+      "loss": 4.3833,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639525460195249e-05,
+      "loss": 4.37,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638686865444197e-05,
+      "loss": 4.3683,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637849908573518e-05,
+      "loss": 4.3696,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637011313822466e-05,
+      "loss": 4.379,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636172719071414e-05,
+      "loss": 4.3738,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635334124320362e-05,
+      "loss": 4.3712,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63449552956931e-05,
+      "loss": 4.3744,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633656934818258e-05,
+      "loss": 4.3739,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632818340067206e-05,
+      "loss": 4.3746,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631979745316154e-05,
+      "loss": 4.362,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631142788445475e-05,
+      "loss": 4.38,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630304193694423e-05,
+      "loss": 4.367,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629467236823744e-05,
+      "loss": 4.3644,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628628642072692e-05,
+      "loss": 4.3667,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627791685202013e-05,
+      "loss": 4.3679,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626953090450961e-05,
+      "loss": 4.3644,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626114495699909e-05,
+      "loss": 4.3539,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625275900948857e-05,
+      "loss": 4.3578,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.342503547668457,
+      "eval_runtime": 298.1181,
+      "eval_samples_per_second": 1279.999,
+      "eval_steps_per_second": 40.001,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624437306197805e-05,
+      "loss": 4.3592,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623598711446753e-05,
+      "loss": 4.3472,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622761754576074e-05,
+      "loss": 4.3747,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621923159825022e-05,
+      "loss": 4.3666,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62108456507397e-05,
+      "loss": 4.3641,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620245970322919e-05,
+      "loss": 4.3495,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619407375571867e-05,
+      "loss": 4.362,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618568780820814e-05,
+      "loss": 4.3419,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617730186069762e-05,
+      "loss": 4.3643,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61689159131871e-05,
+      "loss": 4.358,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616052996567658e-05,
+      "loss": 4.3552,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615214401816606e-05,
+      "loss": 4.3657,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614377444945927e-05,
+      "loss": 4.3444,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613538850194875e-05,
+      "loss": 4.3513,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612700255443823e-05,
+      "loss": 4.3387,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611861660692771e-05,
+      "loss": 4.3485,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611023065941719e-05,
+      "loss": 4.3456,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610184471190667e-05,
+      "loss": 4.3461,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093458764396156e-05,
+      "loss": 4.3465,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085072816885636e-05,
+      "loss": 4.3666,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076703248178845e-05,
+      "loss": 4.3496,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068333679472054e-05,
+      "loss": 4.3532,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059947731961534e-05,
+      "loss": 4.3465,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051561784451014e-05,
+      "loss": 4.3576,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043175836940494e-05,
+      "loss": 4.3388,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6034789889429974e-05,
+      "loss": 4.3472,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602642032072318e-05,
+      "loss": 4.3365,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601803437321266e-05,
+      "loss": 4.3359,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600964842570214e-05,
+      "loss": 4.3359,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600126247819162e-05,
+      "loss": 4.3399,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599287653068111e-05,
+      "loss": 4.338,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598449058317059e-05,
+      "loss": 4.3501,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59761210144638e-05,
+      "loss": 4.3405,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596773506695328e-05,
+      "loss": 4.3428,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595934911944276e-05,
+      "loss": 4.3365,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595096317193224e-05,
+      "loss": 4.3464,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594259360322545e-05,
+      "loss": 4.325,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593420765571493e-05,
+      "loss": 4.3384,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592582170820441e-05,
+      "loss": 4.3175,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591743576069389e-05,
+      "loss": 4.3357,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590904981318337e-05,
+      "loss": 4.3286,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590066386567285e-05,
+      "loss": 4.3412,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589227791816233e-05,
+      "loss": 4.3221,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588389197065181e-05,
+      "loss": 4.3327,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587552240194502e-05,
+      "loss": 4.3348,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867136454434496e-05,
+      "loss": 4.3397,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858750506923976e-05,
+      "loss": 4.3336,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850364559413456e-05,
+      "loss": 4.3214,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584199499070667e-05,
+      "loss": 4.3121,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833609043196145e-05,
+      "loss": 4.3359,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825223095685625e-05,
+      "loss": 4.3312,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816837148175105e-05,
+      "loss": 4.3237,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808451200664585e-05,
+      "loss": 4.3193,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800065253154065e-05,
+      "loss": 4.3221,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791679305643545e-05,
+      "loss": 4.3066,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578330973693676e-05,
+      "loss": 4.3289,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577492378942624e-05,
+      "loss": 4.3062,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576653784191572e-05,
+      "loss": 4.3189,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57581518944052e-05,
+      "loss": 4.3261,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574976594689468e-05,
+      "loss": 4.3143,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574137999938416e-05,
+      "loss": 4.3102,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573299405187364e-05,
+      "loss": 4.3161,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572462448316685e-05,
+      "loss": 4.2996,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571623853565633e-05,
+      "loss": 4.3082,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570785258814581e-05,
+      "loss": 4.3223,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569948301943902e-05,
+      "loss": 4.3107,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56910970719285e-05,
+      "loss": 4.3038,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568271112441798e-05,
+      "loss": 4.3013,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674325176907465e-05,
+      "loss": 4.3101,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5665939229396945e-05,
+      "loss": 4.3164,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657553281886425e-05,
+      "loss": 4.3175,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649167334375905e-05,
+      "loss": 4.3107,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640781386865385e-05,
+      "loss": 4.3187,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632411818158594e-05,
+      "loss": 4.3197,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624025870648074e-05,
+      "loss": 4.3047,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615639923137554e-05,
+      "loss": 4.3033,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607253975627034e-05,
+      "loss": 4.3119,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559886802811651e-05,
+      "loss": 4.3076,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559049845940972e-05,
+      "loss": 4.3053,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55821125118992e-05,
+      "loss": 4.3035,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557372656438868e-05,
+      "loss": 4.3097,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556534061687816e-05,
+      "loss": 4.3153,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555695466936764e-05,
+      "loss": 4.3059,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554856872185712e-05,
+      "loss": 4.2925,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55401827743466e-05,
+      "loss": 4.2923,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553179682683608e-05,
+      "loss": 4.3013,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552344363693303e-05,
+      "loss": 4.31,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551505768942251e-05,
+      "loss": 4.3043,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550668812071572e-05,
+      "loss": 4.2941,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54983021732052e-05,
+      "loss": 4.3075,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548991622569468e-05,
+      "loss": 4.2979,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481530278184157e-05,
+      "loss": 4.295,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473144330673637e-05,
+      "loss": 4.2945,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464758383163116e-05,
+      "loss": 4.2995,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456372435652596e-05,
+      "loss": 4.3096,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5447986488142076e-05,
+      "loss": 4.2919,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439600540631556e-05,
+      "loss": 4.2827,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431214593121036e-05,
+      "loss": 4.2984,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422828645610516e-05,
+      "loss": 4.2925,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414459076903725e-05,
+      "loss": 4.3076,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406073129393205e-05,
+      "loss": 4.2998,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397703560686414e-05,
+      "loss": 4.2894,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389317613175894e-05,
+      "loss": 4.2832,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380931665665374e-05,
+      "loss": 4.2982,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372545718154854e-05,
+      "loss": 4.2782,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536415977064434e-05,
+      "loss": 4.2893,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535579020193755e-05,
+      "loss": 4.2833,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534740425442703e-05,
+      "loss": 4.2918,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533901830691651e-05,
+      "loss": 4.2762,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533063235940599e-05,
+      "loss": 4.2965,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532224641189547e-05,
+      "loss": 4.281,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531386046438495e-05,
+      "loss": 4.2794,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530547451687443e-05,
+      "loss": 4.2924,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529708856936391e-05,
+      "loss": 4.2777,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528870262185339e-05,
+      "loss": 4.3033,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52803330531466e-05,
+      "loss": 4.2913,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5271979863243544e-05,
+      "loss": 4.2878,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263593915733024e-05,
+      "loss": 4.281,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255207968222504e-05,
+      "loss": 4.2841,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246822020711984e-05,
+      "loss": 4.2834,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238436073201464e-05,
+      "loss": 4.2791,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230050125690944e-05,
+      "loss": 4.2898,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221664178180424e-05,
+      "loss": 4.2839,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213278230669904e-05,
+      "loss": 4.279,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5204892283159384e-05,
+      "loss": 4.2791,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196506335648863e-05,
+      "loss": 4.2798,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188120388138343e-05,
+      "loss": 4.2724,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517975081943155e-05,
+      "loss": 4.2885,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517136487192103e-05,
+      "loss": 4.2678,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516297892441051e-05,
+      "loss": 4.2882,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515459297689999e-05,
+      "loss": 4.2831,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514620702938947e-05,
+      "loss": 4.283,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513782108187895e-05,
+      "loss": 4.273,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512943513436843e-05,
+      "loss": 4.2771,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512104918685791e-05,
+      "loss": 4.2861,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511267961815112e-05,
+      "loss": 4.2837,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510431004944434e-05,
+      "loss": 4.2745,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509592410193382e-05,
+      "loss": 4.2807,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508753815442329e-05,
+      "loss": 4.2797,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507915220691277e-05,
+      "loss": 4.2845,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507076625940225e-05,
+      "loss": 4.2704,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506238031189173e-05,
+      "loss": 4.286,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505399436438121e-05,
+      "loss": 4.2771,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50456084168707e-05,
+      "loss": 4.2736,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037238848163906e-05,
+      "loss": 4.2738,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028852900653386e-05,
+      "loss": 4.2792,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020466953142866e-05,
+      "loss": 4.2712,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012081005632346e-05,
+      "loss": 4.2591,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003711436925555e-05,
+      "loss": 4.2754,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.260072708129883,
+      "eval_runtime": 298.381,
+      "eval_samples_per_second": 1278.872,
+      "eval_steps_per_second": 39.966,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995325489415035e-05,
+      "loss": 4.2627,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.2562,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.2819,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.2767,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.2746,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.266,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.2706,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.2511,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.2751,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919851961820354e-05,
+      "loss": 4.273,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911466014309834e-05,
+      "loss": 4.2664,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903080066799314e-05,
+      "loss": 4.2761,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894694119288794e-05,
+      "loss": 4.2578,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886308171778274e-05,
+      "loss": 4.2638,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877922224267754e-05,
+      "loss": 4.2565,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869536276757234e-05,
+      "loss": 4.2557,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861150329246714e-05,
+      "loss": 4.26,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4852764381736194e-05,
+      "loss": 4.2591,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844378434225674e-05,
+      "loss": 4.2619,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483600886551888e-05,
+      "loss": 4.2778,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482762291800836e-05,
+      "loss": 4.266,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481923697049785e-05,
+      "loss": 4.2699,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481085102298733e-05,
+      "loss": 4.2613,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480246507547681e-05,
+      "loss": 4.2678,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479409550677002e-05,
+      "loss": 4.2555,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2618,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477732361174898e-05,
+      "loss": 4.2575,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476893766423846e-05,
+      "loss": 4.249,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476055171672794e-05,
+      "loss": 4.25,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475216576921742e-05,
+      "loss": 4.2565,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474377982170689e-05,
+      "loss": 4.2551,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473539387419637e-05,
+      "loss": 4.2658,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.258,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471863835797907e-05,
+      "loss": 4.2598,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710268789272283e-05,
+      "loss": 4.2511,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701882841761763e-05,
+      "loss": 4.2657,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693496894251243e-05,
+      "loss": 4.2465,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685110946740717e-05,
+      "loss": 4.2523,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676724999230197e-05,
+      "loss": 4.2375,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466835543052341e-05,
+      "loss": 4.2521,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465996948301289e-05,
+      "loss": 4.2493,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651583535502366e-05,
+      "loss": 4.2576,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643197587991846e-05,
+      "loss": 4.2424,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463482801928506e-05,
+      "loss": 4.2526,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462644207177454e-05,
+      "loss": 4.2548,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461805612426402e-05,
+      "loss": 4.2559,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46096701767535e-05,
+      "loss": 4.2528,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460128422924298e-05,
+      "loss": 4.2404,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459289828173246e-05,
+      "loss": 4.2309,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458452871302567e-05,
+      "loss": 4.2585,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457614276551515e-05,
+      "loss": 4.2501,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567773196808366e-05,
+      "loss": 4.2442,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455938724929784e-05,
+      "loss": 4.2429,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455100130178732e-05,
+      "loss": 4.2431,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45426153542768e-05,
+      "loss": 4.2278,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.2449,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452584345925576e-05,
+      "loss": 4.2307,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451745751174524e-05,
+      "loss": 4.2373,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450907156423472e-05,
+      "loss": 4.2499,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500685616724206e-05,
+      "loss": 4.2367,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492299669213686e-05,
+      "loss": 4.2306,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483913721703166e-05,
+      "loss": 4.2443,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475527774192646e-05,
+      "loss": 4.2227,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467158205485855e-05,
+      "loss": 4.2324,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458772257975335e-05,
+      "loss": 4.2419,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450386310464815e-05,
+      "loss": 4.2381,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442000362954295e-05,
+      "loss": 4.2281,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4433614415443775e-05,
+      "loss": 4.2274,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425228467933254e-05,
+      "loss": 4.2292,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416858899226464e-05,
+      "loss": 4.2409,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408472951715944e-05,
+      "loss": 4.2399,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400087004205423e-05,
+      "loss": 4.2347,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391701056694903e-05,
+      "loss": 4.2425,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438333148798812e-05,
+      "loss": 4.2455,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43749455404776e-05,
+      "loss": 4.2318,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436655959296708e-05,
+      "loss": 4.223,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435817364545655e-05,
+      "loss": 4.2418,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434978769794603e-05,
+      "loss": 4.2345,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434140175043551e-05,
+      "loss": 4.2273,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433301580292499e-05,
+      "loss": 4.2304,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432462985541447e-05,
+      "loss": 4.2344,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431627666551142e-05,
+      "loss": 4.2439,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43078907180009e-05,
+      "loss": 4.228,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429950477049038e-05,
+      "loss": 4.2178,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429111882297986e-05,
+      "loss": 4.2267,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428273287546934e-05,
+      "loss": 4.2212,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427434692795882e-05,
+      "loss": 4.2342,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42659609804483e-05,
+      "loss": 4.2321,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257591411741506e-05,
+      "loss": 4.2185,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249205464230986e-05,
+      "loss": 4.2352,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240819516720466e-05,
+      "loss": 4.2247,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232449948013675e-05,
+      "loss": 4.2258,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224064000503155e-05,
+      "loss": 4.2188,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215678052992635e-05,
+      "loss": 4.2266,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207292105482115e-05,
+      "loss": 4.2379,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198906157971595e-05,
+      "loss": 4.2257,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419052021046108e-05,
+      "loss": 4.208,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418213426295056e-05,
+      "loss": 4.2225,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417374831544004e-05,
+      "loss": 4.2231,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416536236792952e-05,
+      "loss": 4.2308,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4156976420419e-05,
+      "loss": 4.2328,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414859047290848e-05,
+      "loss": 4.2188,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414022090420169e-05,
+      "loss": 4.2146,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413183495669117e-05,
+      "loss": 4.2263,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412346538798438e-05,
+      "loss": 4.2048,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411507944047386e-05,
+      "loss": 4.2204,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410669349296334e-05,
+      "loss": 4.2121,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409830754545282e-05,
+      "loss": 4.2216,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899215979423e-05,
+      "loss": 4.208,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408153565043178e-05,
+      "loss": 4.2258,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407314970292126e-05,
+      "loss": 4.2101,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406476375541074e-05,
+      "loss": 4.2098,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056394186703955e-05,
+      "loss": 4.2245,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048008239193435e-05,
+      "loss": 4.213,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039622291682915e-05,
+      "loss": 4.2319,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403123634417239e-05,
+      "loss": 4.2219,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4022866775465604e-05,
+      "loss": 4.2137,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014480827955084e-05,
+      "loss": 4.2205,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400609488044456e-05,
+      "loss": 4.2084,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399770893293404e-05,
+      "loss": 4.2191,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398932298542352e-05,
+      "loss": 4.2095,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3980937037913e-05,
+      "loss": 4.2245,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972551090402484e-05,
+      "loss": 4.2123,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964165142891964e-05,
+      "loss": 4.2102,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3955779195381444e-05,
+      "loss": 4.2108,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947393247870924e-05,
+      "loss": 4.2144,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393902367916413e-05,
+      "loss": 4.2037,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393063773165361e-05,
+      "loss": 4.2194,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392225178414309e-05,
+      "loss": 4.2018,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391386583663257e-05,
+      "loss": 4.2224,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390547988912205e-05,
+      "loss": 4.2154,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389709394161153e-05,
+      "loss": 4.2156,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388870799410101e-05,
+      "loss": 4.2065,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388033842539422e-05,
+      "loss": 4.2081,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38719524778837e-05,
+      "loss": 4.2221,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386356653037318e-05,
+      "loss": 4.2143,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385518058286267e-05,
+      "loss": 4.2058,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384679463535215e-05,
+      "loss": 4.217,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383840868784163e-05,
+      "loss": 4.2139,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38300227403311e-05,
+      "loss": 4.2205,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382165317162432e-05,
+      "loss": 4.2068,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38132672241138e-05,
+      "loss": 4.2188,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380488127660328e-05,
+      "loss": 4.2113,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796511707896486e-05,
+      "loss": 4.2073,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788125760385966e-05,
+      "loss": 4.2104,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779739812875446e-05,
+      "loss": 4.2116,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377135386536492e-05,
+      "loss": 4.2101,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3762967917854406e-05,
+      "loss": 4.1969,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3754581970343886e-05,
+      "loss": 4.2103,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.204163551330566,
+      "eval_runtime": 286.8663,
+      "eval_samples_per_second": 1330.205,
+      "eval_steps_per_second": 41.57,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3746196022833366e-05,
+      "loss": 4.2054,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3737810075322846e-05,
+      "loss": 4.1946,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3729440506616055e-05,
+      "loss": 4.2167,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721054559105535e-05,
+      "loss": 4.2142,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371268499039875e-05,
+      "loss": 4.2089,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3704299042888224e-05,
+      "loss": 4.2034,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3695913095377704e-05,
+      "loss": 4.2046,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3687527147867184e-05,
+      "loss": 4.1919,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3679141200356664e-05,
+      "loss": 4.2085,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670755252846144e-05,
+      "loss": 4.213,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662369305335624e-05,
+      "loss": 4.2047,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3653983357825104e-05,
+      "loss": 4.2098,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364563016792205e-05,
+      "loss": 4.1978,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363724422041153e-05,
+      "loss": 4.2022,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362885827290101e-05,
+      "loss": 4.1921,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362047232539049e-05,
+      "loss": 4.1924,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361208637787997e-05,
+      "loss": 4.1973,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360370043036945e-05,
+      "loss": 4.1974,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359531448285893e-05,
+      "loss": 4.1999,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358692853534841e-05,
+      "loss": 4.2131,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357855896664162e-05,
+      "loss": 4.2101,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35701730191311e-05,
+      "loss": 4.2091,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356178707162058e-05,
+      "loss": 4.1964,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355340112411006e-05,
+      "loss": 4.2088,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354503155540327e-05,
+      "loss": 4.1915,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353666198669648e-05,
+      "loss": 4.2045,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352827603918596e-05,
+      "loss": 4.1946,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351989009167544e-05,
+      "loss": 4.185,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351150414416492e-05,
+      "loss": 4.193,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35031181966544e-05,
+      "loss": 4.1921,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349473224914388e-05,
+      "loss": 4.1958,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348634630163336e-05,
+      "loss": 4.201,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347796035412284e-05,
+      "loss": 4.2021,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346959078541605e-05,
+      "loss": 4.1996,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346122121670926e-05,
+      "loss": 4.1907,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345283526919875e-05,
+      "loss": 4.2019,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344444932168823e-05,
+      "loss": 4.1891,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436079752981436e-05,
+      "loss": 4.1886,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427693805470916e-05,
+      "loss": 4.1836,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419307857960396e-05,
+      "loss": 4.189,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3410921910449876e-05,
+      "loss": 4.1929,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402535962939356e-05,
+      "loss": 4.1968,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394150015428836e-05,
+      "loss": 4.185,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385764067918316e-05,
+      "loss": 4.1914,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377378120407796e-05,
+      "loss": 4.1962,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3368992172897276e-05,
+      "loss": 4.1948,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360606225386756e-05,
+      "loss": 4.1974,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352236656679965e-05,
+      "loss": 4.1815,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3343850709169445e-05,
+      "loss": 4.1688,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333546476165893e-05,
+      "loss": 4.2017,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332707881414841e-05,
+      "loss": 4.1935,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3318692866637885e-05,
+      "loss": 4.1887,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.1889,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330193735042058e-05,
+      "loss": 4.1785,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3293551402910054e-05,
+      "loss": 4.1735,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285165455399534e-05,
+      "loss": 4.1846,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327679588669275e-05,
+      "loss": 4.1731,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326840993918223e-05,
+      "loss": 4.181,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32600239916717e-05,
+      "loss": 4.1932,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325163804416118e-05,
+      "loss": 4.1787,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324325209665067e-05,
+      "loss": 4.1729,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323486614914015e-05,
+      "loss": 4.1889,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322648020162963e-05,
+      "loss": 4.1684,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321809425411911e-05,
+      "loss": 4.1728,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320972468541232e-05,
+      "loss": 4.1854,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32013387379018e-05,
+      "loss": 4.1828,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319295279039128e-05,
+      "loss": 4.1728,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318456684288076e-05,
+      "loss": 4.1706,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317619727417397e-05,
+      "loss": 4.171,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316781132666345e-05,
+      "loss": 4.1816,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315942537915293e-05,
+      "loss": 4.1856,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315103943164241e-05,
+      "loss": 4.1757,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314265348413189e-05,
+      "loss": 4.1846,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31342839154251e-05,
+      "loss": 4.1883,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312589796791458e-05,
+      "loss": 4.1835,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311754477801153e-05,
+      "loss": 4.1638,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109158830501e-05,
+      "loss": 4.1828,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310077288299048e-05,
+      "loss": 4.1832,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309238693547996e-05,
+      "loss": 4.1686,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308400098796944e-05,
+      "loss": 4.1814,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307561504045892e-05,
+      "loss": 4.1749,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30672290929484e-05,
+      "loss": 4.188,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305884314543788e-05,
+      "loss": 4.1764,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305045719792736e-05,
+      "loss": 4.1573,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304207125041684e-05,
+      "loss": 4.1741,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303368530290632e-05,
+      "loss": 4.1688,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025315734199536e-05,
+      "loss": 4.177,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3016929786689016e-05,
+      "loss": 4.1758,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008543839178496e-05,
+      "loss": 4.1691,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000157891667976e-05,
+      "loss": 4.1765,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991771944157456e-05,
+      "loss": 4.1737,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983385996646936e-05,
+      "loss": 4.1673,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975016427940145e-05,
+      "loss": 4.1692,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966630480429625e-05,
+      "loss": 4.1754,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958244532919105e-05,
+      "loss": 4.1842,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949858585408585e-05,
+      "loss": 4.1699,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941472637898065e-05,
+      "loss": 4.1563,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293308669038754e-05,
+      "loss": 4.1693,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924700742877025e-05,
+      "loss": 4.17,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916314795366505e-05,
+      "loss": 4.1787,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2907945226659714e-05,
+      "loss": 4.18,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2899559279149194e-05,
+      "loss": 4.1623,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2891173331638674e-05,
+      "loss": 4.1619,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2882787384128154e-05,
+      "loss": 4.1742,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287441781542136e-05,
+      "loss": 4.152,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286603186791084e-05,
+      "loss": 4.1697,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285764592040032e-05,
+      "loss": 4.1556,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28492599728898e-05,
+      "loss": 4.1705,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284090678298675e-05,
+      "loss": 4.1567,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283252083547623e-05,
+      "loss": 4.1723,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282413488796571e-05,
+      "loss": 4.1625,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281574894045519e-05,
+      "loss": 4.155,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280736299294467e-05,
+      "loss": 4.1719,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279897704543415e-05,
+      "loss": 4.1585,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279059109792363e-05,
+      "loss": 4.1813,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278220515041311e-05,
+      "loss": 4.1751,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277385196051005e-05,
+      "loss": 4.1609,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276546601299953e-05,
+      "loss": 4.1709,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275708006548901e-05,
+      "loss": 4.1577,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748694117978486e-05,
+      "loss": 4.166,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740308170467966e-05,
+      "loss": 4.162,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2731922222957446e-05,
+      "loss": 4.17,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272353627544693e-05,
+      "loss": 4.162,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271515032793641e-05,
+      "loss": 4.1586,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270676438042589e-05,
+      "loss": 4.1632,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26983948117191e-05,
+      "loss": 4.162,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269000886420858e-05,
+      "loss": 4.152,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268163929550179e-05,
+      "loss": 4.1701,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267325334799127e-05,
+      "loss": 4.1482,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266486740048075e-05,
+      "loss": 4.1778,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265648145297023e-05,
+      "loss": 4.1583,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264809550545971e-05,
+      "loss": 4.1686,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263970955794919e-05,
+      "loss": 4.1545,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263132361043867e-05,
+      "loss": 4.1615,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262293766292815e-05,
+      "loss": 4.1693,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261455171541763e-05,
+      "loss": 4.162,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260616576790712e-05,
+      "loss": 4.1581,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25977798203966e-05,
+      "loss": 4.1685,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258939387288608e-05,
+      "loss": 4.1631,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581024304179286e-05,
+      "loss": 4.17,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2572638356668766e-05,
+      "loss": 4.1537,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564252409158246e-05,
+      "loss": 4.1677,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2555866461647726e-05,
+      "loss": 4.1618,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547496892940935e-05,
+      "loss": 4.1612,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539110945430415e-05,
+      "loss": 4.163,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2530724997919895e-05,
+      "loss": 4.1604,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522339050409375e-05,
+      "loss": 4.1666,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2513953102898855e-05,
+      "loss": 4.1397,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250558353419207e-05,
+      "loss": 4.167,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.163326263427734,
+      "eval_runtime": 290.1644,
+      "eval_samples_per_second": 1315.085,
+      "eval_steps_per_second": 41.097,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249719758668155e-05,
+      "loss": 4.155,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488811639171024e-05,
+      "loss": 4.1456,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480425691660504e-05,
+      "loss": 4.1651,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472039744149984e-05,
+      "loss": 4.1679,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2463653796639464e-05,
+      "loss": 4.1587,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455267849128943e-05,
+      "loss": 4.1519,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2446881901618423e-05,
+      "loss": 4.157,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24384959541079e-05,
+      "loss": 4.1447,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243011000659738e-05,
+      "loss": 4.1586,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242172405908686e-05,
+      "loss": 4.1615,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241333811157634e-05,
+      "loss": 4.1618,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240495216406582e-05,
+      "loss": 4.1592,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239658259535903e-05,
+      "loss": 4.1539,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238819664784852e-05,
+      "loss": 4.152,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379810700338e-05,
+      "loss": 4.146,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237142475282748e-05,
+      "loss": 4.1407,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236303880531696e-05,
+      "loss": 4.1512,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235465285780644e-05,
+      "loss": 4.1478,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234626691029591e-05,
+      "loss": 4.1524,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233788096278539e-05,
+      "loss": 4.1644,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232949501527487e-05,
+      "loss": 4.1667,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232110906776435e-05,
+      "loss": 4.158,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231272312025383e-05,
+      "loss": 4.1501,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230433717274331e-05,
+      "loss": 4.1618,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229598398284026e-05,
+      "loss": 4.145,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228759803532974e-05,
+      "loss": 4.1597,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227921208781922e-05,
+      "loss": 4.1482,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22708261403087e-05,
+      "loss": 4.1359,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226245657160191e-05,
+      "loss": 4.1502,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254070624091386e-05,
+      "loss": 4.1462,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245684676580866e-05,
+      "loss": 4.1472,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237298729070346e-05,
+      "loss": 4.1583,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2228912781559826e-05,
+      "loss": 4.1543,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220526834049306e-05,
+      "loss": 4.1533,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212157265342515e-05,
+      "loss": 4.1431,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203771317831995e-05,
+      "loss": 4.1619,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195385370321475e-05,
+      "loss": 4.1438,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218701580161469e-05,
+      "loss": 4.1377,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217862985410417e-05,
+      "loss": 4.1428,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217024390659365e-05,
+      "loss": 4.1373,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216185795908313e-05,
+      "loss": 4.1481,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215347201157261e-05,
+      "loss": 4.1547,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214508606406209e-05,
+      "loss": 4.1372,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21367164953553e-05,
+      "loss": 4.1453,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212833054784478e-05,
+      "loss": 4.1528,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211994460033426e-05,
+      "loss": 4.1538,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211155865282374e-05,
+      "loss": 4.1519,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210317270531322e-05,
+      "loss": 4.1399,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20947867578027e-05,
+      "loss": 4.115,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208641718909591e-05,
+      "loss": 4.1595,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078031241585395e-05,
+      "loss": 4.1496,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2069645294074875e-05,
+      "loss": 4.1478,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061259346564355e-05,
+      "loss": 4.1412,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052873399053835e-05,
+      "loss": 4.1341,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044487451543315e-05,
+      "loss": 4.1299,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036117882836524e-05,
+      "loss": 4.136,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027731935326004e-05,
+      "loss": 4.1321,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019345987815484e-05,
+      "loss": 4.1376,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2010960040304964e-05,
+      "loss": 4.15,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2002574092794444e-05,
+      "loss": 4.1324,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994188145283924e-05,
+      "loss": 4.1283,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198581857657713e-05,
+      "loss": 4.1446,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197743262906661e-05,
+      "loss": 4.1278,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196904668155609e-05,
+      "loss": 4.13,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196066073404557e-05,
+      "loss": 4.1368,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195227478653505e-05,
+      "loss": 4.1402,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194388883902453e-05,
+      "loss": 4.1275,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193551927031775e-05,
+      "loss": 4.1262,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192713332280722e-05,
+      "loss": 4.1258,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19187473752967e-05,
+      "loss": 4.1428,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191036142778618e-05,
+      "loss": 4.1385,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190197548027566e-05,
+      "loss": 4.1355,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189358953276514e-05,
+      "loss": 4.1358,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188520358525462e-05,
+      "loss": 4.1452,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18768176377441e-05,
+      "loss": 4.1422,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186843169023358e-05,
+      "loss": 4.1262,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186004574272306e-05,
+      "loss": 4.1371,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185165979521254e-05,
+      "loss": 4.1439,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184327384770203e-05,
+      "loss": 4.123,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183490427899524e-05,
+      "loss": 4.1374,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826534710288446e-05,
+      "loss": 4.1344,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818148762777926e-05,
+      "loss": 4.143,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809762815267406e-05,
+      "loss": 4.1357,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801376867756886e-05,
+      "loss": 4.1175,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1792990920246366e-05,
+      "loss": 4.1293,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784604972735846e-05,
+      "loss": 4.1285,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776219025225326e-05,
+      "loss": 4.1382,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767833077714806e-05,
+      "loss": 4.1311,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1759463509008015e-05,
+      "loss": 4.1262,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751077561497495e-05,
+      "loss": 4.1326,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174269161398698e-05,
+      "loss": 4.1366,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173430566647646e-05,
+      "loss": 4.1216,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172593609776967e-05,
+      "loss": 4.1256,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171755015025915e-05,
+      "loss": 4.1364,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170918058155236e-05,
+      "loss": 4.1408,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170079463404184e-05,
+      "loss": 4.1285,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169240868653132e-05,
+      "loss": 4.1179,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16840227390208e-05,
+      "loss": 4.1261,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167563679151028e-05,
+      "loss": 4.1265,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166726722280349e-05,
+      "loss": 4.1368,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165888127529297e-05,
+      "loss": 4.1397,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165049532778245e-05,
+      "loss": 4.1247,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642125759075664e-05,
+      "loss": 4.1228,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633739811565144e-05,
+      "loss": 4.1303,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625353864054624e-05,
+      "loss": 4.1123,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1616967916544104e-05,
+      "loss": 4.1302,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1608581969033584e-05,
+      "loss": 4.1109,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160019602152306e-05,
+      "loss": 4.1309,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159181007401254e-05,
+      "loss": 4.1179,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158342412650202e-05,
+      "loss": 4.1284,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15750381789915e-05,
+      "loss": 4.1225,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156665223148098e-05,
+      "loss": 4.1198,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558282662774186e-05,
+      "loss": 4.1293,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154989671526367e-05,
+      "loss": 4.118,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154151076775315e-05,
+      "loss": 4.1421,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153312482024263e-05,
+      "loss": 4.1342,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152473887273211e-05,
+      "loss": 4.1218,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151636930402532e-05,
+      "loss": 4.1324,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15079833565148e-05,
+      "loss": 4.1163,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149959740900428e-05,
+      "loss": 4.1256,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149121146149376e-05,
+      "loss": 4.1249,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148282551398324e-05,
+      "loss": 4.1268,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147445594527645e-05,
+      "loss": 4.1219,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146606999776593e-05,
+      "loss": 4.1204,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145768405025541e-05,
+      "loss": 4.1258,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144929810274489e-05,
+      "loss": 4.1217,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144091215523437e-05,
+      "loss": 4.116,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143252620772386e-05,
+      "loss": 4.1285,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142414026021334e-05,
+      "loss": 4.1077,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415770691506546e-05,
+      "loss": 4.1372,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407384743996026e-05,
+      "loss": 4.1236,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1398998796485506e-05,
+      "loss": 4.1291,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390612848974986e-05,
+      "loss": 4.1106,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382226901464466e-05,
+      "loss": 4.1259,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1373857332757675e-05,
+      "loss": 4.129,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1365471385247155e-05,
+      "loss": 4.1259,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357085437736635e-05,
+      "loss": 4.1175,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1348699490226115e-05,
+      "loss": 4.1306,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340313542715595e-05,
+      "loss": 4.1228,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1331927595205075e-05,
+      "loss": 4.1338,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323541647694555e-05,
+      "loss": 4.1134,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315155700184035e-05,
+      "loss": 4.1307,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306786131477244e-05,
+      "loss": 4.1215,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298400183966724e-05,
+      "loss": 4.1216,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129003061525994e-05,
+      "loss": 4.1235,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128164466774942e-05,
+      "loss": 4.1222,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127325872023889e-05,
+      "loss": 4.1244,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126487277272837e-05,
+      "loss": 4.1067,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125650320402159e-05,
+      "loss": 4.1277,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.133300304412842,
+      "eval_runtime": 296.1158,
+      "eval_samples_per_second": 1288.655,
+      "eval_steps_per_second": 40.271,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124811725651106e-05,
+      "loss": 4.116,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123973130900054e-05,
+      "loss": 4.1084,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123134536149003e-05,
+      "loss": 4.124,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122295941397951e-05,
+      "loss": 4.1293,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121457346646899e-05,
+      "loss": 4.1201,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120618751895847e-05,
+      "loss": 4.1168,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119780157144795e-05,
+      "loss": 4.1226,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118941562393743e-05,
+      "loss": 4.1073,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118102967642691e-05,
+      "loss": 4.1209,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117264372891639e-05,
+      "loss": 4.1239,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116425778140587e-05,
+      "loss": 4.1224,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115588821269908e-05,
+      "loss": 4.124,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114751864399229e-05,
+      "loss": 4.1142,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113913269648177e-05,
+      "loss": 4.1145,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1130746748971247e-05,
+      "loss": 4.1116,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122360801460727e-05,
+      "loss": 4.1026,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111397485395021e-05,
+      "loss": 4.1149,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110560528524342e-05,
+      "loss": 4.11,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10972193377329e-05,
+      "loss": 4.1152,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108883339022238e-05,
+      "loss": 4.1248,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108044744271186e-05,
+      "loss": 4.1294,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107206149520134e-05,
+      "loss": 4.1255,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106367554769082e-05,
+      "loss": 4.1114,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10552896001803e-05,
+      "loss": 4.1223,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1046903652669775e-05,
+      "loss": 4.1105,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1038517705159255e-05,
+      "loss": 4.1198,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030131757648735e-05,
+      "loss": 4.1133,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021745810138215e-05,
+      "loss": 4.1003,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101339262023517e-05,
+      "loss": 4.1152,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100500667272465e-05,
+      "loss": 4.1032,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099662072521413e-05,
+      "loss": 4.1106,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098823477770361e-05,
+      "loss": 4.1217,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097984883019308e-05,
+      "loss": 4.115,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097146288268256e-05,
+      "loss": 4.1156,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096307693517204e-05,
+      "loss": 4.1044,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095470736646525e-05,
+      "loss": 4.127,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0946337797758465e-05,
+      "loss": 4.1117,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0937951850247945e-05,
+      "loss": 4.0994,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0929565902737425e-05,
+      "loss": 4.1096,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921179955226905e-05,
+      "loss": 4.1007,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0912794007716385e-05,
+      "loss": 4.1083,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0904408060205865e-05,
+      "loss": 4.114,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896022112695345e-05,
+      "loss": 4.1056,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0887636165184825e-05,
+      "loss": 4.1053,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879250217674305e-05,
+      "loss": 4.1162,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0870864270163784e-05,
+      "loss": 4.1148,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0862478322653264e-05,
+      "loss": 4.1176,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854092375142744e-05,
+      "loss": 4.1011,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0845722806435953e-05,
+      "loss": 4.0783,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837336858925433e-05,
+      "loss": 4.1274,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082896729021864e-05,
+      "loss": 4.1089,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082058134270812e-05,
+      "loss": 4.1133,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08121953951976e-05,
+      "loss": 4.1057,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080380944768708e-05,
+      "loss": 4.0957,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079542350017657e-05,
+      "loss": 4.0974,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078705393146978e-05,
+      "loss": 4.1009,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077866798395926e-05,
+      "loss": 4.0949,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077028203644874e-05,
+      "loss": 4.0991,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076189608893822e-05,
+      "loss": 4.1184,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07535101414277e-05,
+      "loss": 4.0945,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074512419391718e-05,
+      "loss": 4.0908,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073673824640666e-05,
+      "loss": 4.1141,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072835229889614e-05,
+      "loss": 4.0891,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071996635138561e-05,
+      "loss": 4.0937,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071159678267883e-05,
+      "loss": 4.1028,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070321083516831e-05,
+      "loss": 4.1058,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069482488765779e-05,
+      "loss": 4.092,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068643894014727e-05,
+      "loss": 4.0984,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067806937144048e-05,
+      "loss": 4.0865,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066969980273369e-05,
+      "loss": 4.106,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066131385522317e-05,
+      "loss": 4.104,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065292790771265e-05,
+      "loss": 4.0992,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064454196020213e-05,
+      "loss": 4.1034,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063615601269161e-05,
+      "loss": 4.1057,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627770065181085e-05,
+      "loss": 4.111,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619384117670565e-05,
+      "loss": 4.0905,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0610998170160045e-05,
+      "loss": 4.1011,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060262860145326e-05,
+      "loss": 4.1086,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059424265394274e-05,
+      "loss": 4.0846,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058585670643222e-05,
+      "loss": 4.106,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0577487137725436e-05,
+      "loss": 4.0985,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056910119021491e-05,
+      "loss": 4.11,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056071524270439e-05,
+      "loss": 4.1056,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055232929519387e-05,
+      "loss": 4.0861,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0543959726487085e-05,
+      "loss": 4.0858,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053557377897656e-05,
+      "loss": 4.0954,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052718783146604e-05,
+      "loss": 4.1049,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051880188395552e-05,
+      "loss": 4.0935,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0510415936445e-05,
+      "loss": 4.0949,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0502046367738214e-05,
+      "loss": 4.0982,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0493660420227694e-05,
+      "loss": 4.1017,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0485274472717174e-05,
+      "loss": 4.0917,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0476888525206654e-05,
+      "loss": 4.0914,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0468502577696134e-05,
+      "loss": 4.102,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046013300898934e-05,
+      "loss": 4.105,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045174706147882e-05,
+      "loss": 4.0961,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04433611139683e-05,
+      "loss": 4.0841,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043497516645778e-05,
+      "loss": 4.0938,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042658921894726e-05,
+      "loss": 4.0895,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041820327143674e-05,
+      "loss": 4.1057,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040981732392622e-05,
+      "loss": 4.0994,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04014313764157e-05,
+      "loss": 4.0944,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039306180770891e-05,
+      "loss": 4.0912,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03846758601984e-05,
+      "loss": 4.0886,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037628991268788e-05,
+      "loss": 4.083,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036790396517736e-05,
+      "loss": 4.0957,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035956715407803e-05,
+      "loss": 4.0793,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351181206567506e-05,
+      "loss": 4.0994,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342795259056986e-05,
+      "loss": 4.0866,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0334409311546466e-05,
+      "loss": 4.0907,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326023364035946e-05,
+      "loss": 4.0868,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0317637416525426e-05,
+      "loss": 4.0847,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309251469014906e-05,
+      "loss": 4.0977,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0300865521504386e-05,
+      "loss": 4.0869,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0292479573993866e-05,
+      "loss": 4.1084,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028409362648335e-05,
+      "loss": 4.1008,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027572405777656e-05,
+      "loss": 4.0894,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026735448906977e-05,
+      "loss": 4.0963,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025896854155925e-05,
+      "loss": 4.0855,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025058259404873e-05,
+      "loss": 4.0889,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024219664653821e-05,
+      "loss": 4.0968,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023381069902769e-05,
+      "loss": 4.0897,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022542475151717e-05,
+      "loss": 4.0903,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021703880400665e-05,
+      "loss": 4.09,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020865285649613e-05,
+      "loss": 4.0914,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020026690898561e-05,
+      "loss": 4.0917,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019188096147509e-05,
+      "loss": 4.0789,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018349501396457e-05,
+      "loss": 4.094,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017510906645405e-05,
+      "loss": 4.0794,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0166739497747266e-05,
+      "loss": 4.1036,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0158353550236746e-05,
+      "loss": 4.0898,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014996760272622e-05,
+      "loss": 4.0936,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01415816552157e-05,
+      "loss": 4.0809,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013319570770518e-05,
+      "loss": 4.0935,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012480976019466e-05,
+      "loss": 4.0949,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011642381268414e-05,
+      "loss": 4.0919,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010803786517362e-05,
+      "loss": 4.0876,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009966829646683e-05,
+      "loss": 4.0971,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009128234895631e-05,
+      "loss": 4.0892,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008289640144579e-05,
+      "loss": 4.1056,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0074526832739004e-05,
+      "loss": 4.0825,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0066140885228484e-05,
+      "loss": 4.0988,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0057754937717964e-05,
+      "loss": 4.0903,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0049368990207444e-05,
+      "loss": 4.0898,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0040983042696924e-05,
+      "loss": 4.0908,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0032597095186404e-05,
+      "loss": 4.0907,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024211147675883e-05,
+      "loss": 4.0923,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0015825200165363e-05,
+      "loss": 4.0757,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000745563145857e-05,
+      "loss": 4.0937,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.109120845794678,
+      "eval_runtime": 295.692,
+      "eval_samples_per_second": 1290.501,
+      "eval_steps_per_second": 40.329,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999908606275178e-05,
+      "loss": 4.08,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999070011524126e-05,
+      "loss": 4.0782,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998231416773074e-05,
+      "loss": 4.0893,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997392822022022e-05,
+      "loss": 4.1006,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996554227270971e-05,
+      "loss": 4.0897,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995715632519919e-05,
+      "loss": 4.0834,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994877037768867e-05,
+      "loss": 4.0955,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994038443017815e-05,
+      "loss": 4.0741,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993199848266763e-05,
+      "loss": 4.0869,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992361253515711e-05,
+      "loss": 4.0967,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991522658764658e-05,
+      "loss": 4.0895,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990684064013606e-05,
+      "loss": 4.0887,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989847107142928e-05,
+      "loss": 4.0921,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989008512391876e-05,
+      "loss": 4.0785,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988169917640823e-05,
+      "loss": 4.0764,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987331322889771e-05,
+      "loss": 4.0744,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986492728138719e-05,
+      "loss": 4.0821,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985654133387668e-05,
+      "loss": 4.0832,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984815538636616e-05,
+      "loss": 4.0802,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983976943885564e-05,
+      "loss": 4.0919,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983138349134512e-05,
+      "loss": 4.0989,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9822997543834597e-05,
+      "loss": 4.0938,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9814611596324077e-05,
+      "loss": 4.0839,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9806242027617286e-05,
+      "loss": 4.0874,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9797856080106766e-05,
+      "loss": 4.0864,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9789470132596246e-05,
+      "loss": 4.0828,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781084185085726e-05,
+      "loss": 4.085,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9772698237575205e-05,
+      "loss": 4.0751,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9764312290064685e-05,
+      "loss": 4.0823,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9755926342554165e-05,
+      "loss": 4.0751,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9747540395043645e-05,
+      "loss": 4.0775,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973917082633686e-05,
+      "loss": 4.0892,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973078487882634e-05,
+      "loss": 4.0873,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972239893131582e-05,
+      "loss": 4.0854,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9714012983805294e-05,
+      "loss": 4.0783,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970564341509851e-05,
+      "loss": 4.0894,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969725746758799e-05,
+      "loss": 4.0828,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968887152007747e-05,
+      "loss": 4.0728,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968048557256694e-05,
+      "loss": 4.0808,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967209962505642e-05,
+      "loss": 4.068,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966373005634964e-05,
+      "loss": 4.077,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965534410883912e-05,
+      "loss": 4.0859,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96469581613286e-05,
+      "loss": 4.0738,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963857221381808e-05,
+      "loss": 4.0776,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963018626630756e-05,
+      "loss": 4.0903,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962180031879704e-05,
+      "loss": 4.0832,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961341437128652e-05,
+      "loss": 4.0855,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605028423776e-05,
+      "loss": 4.076,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959664247626548e-05,
+      "loss": 4.0509,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958825652875496e-05,
+      "loss": 4.0919,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957987058124444e-05,
+      "loss": 4.0798,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957150101253765e-05,
+      "loss": 4.0826,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956311506502713e-05,
+      "loss": 4.0797,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955472911751661e-05,
+      "loss": 4.0683,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954634317000609e-05,
+      "loss": 4.064,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95379736012993e-05,
+      "loss": 4.0736,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952958765378878e-05,
+      "loss": 4.0658,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9521201706278263e-05,
+      "loss": 4.0673,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951281575876774e-05,
+      "loss": 4.0888,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950442981125722e-05,
+      "loss": 4.0678,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949606024255043e-05,
+      "loss": 4.0612,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948767429503991e-05,
+      "loss": 4.0875,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947928834752939e-05,
+      "loss": 4.0557,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947090240001887e-05,
+      "loss": 4.0628,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946251645250835e-05,
+      "loss": 4.0758,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945413050499783e-05,
+      "loss": 4.0768,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445744557487305e-05,
+      "loss": 4.0627,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943737498878052e-05,
+      "loss": 4.0671,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942898904127e-05,
+      "loss": 4.0601,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942060309375948e-05,
+      "loss": 4.0759,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941221714624896e-05,
+      "loss": 4.0726,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940383119873844e-05,
+      "loss": 4.0708,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939544525122792e-05,
+      "loss": 4.0734,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93870593037174e-05,
+      "loss": 4.0754,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937867335620688e-05,
+      "loss": 4.0855,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937030378750009e-05,
+      "loss": 4.061,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936191783998957e-05,
+      "loss": 4.0752,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935353189247905e-05,
+      "loss": 4.0789,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934516232377226e-05,
+      "loss": 4.0526,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933677637626174e-05,
+      "loss": 4.0751,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932839042875122e-05,
+      "loss": 4.0734,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93200044812407e-05,
+      "loss": 4.0834,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9311618533730186e-05,
+      "loss": 4.0734,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303232586219666e-05,
+      "loss": 4.0605,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9294846638709146e-05,
+      "loss": 4.0528,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9286460691198626e-05,
+      "loss": 4.0668,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278091122491835e-05,
+      "loss": 4.0753,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9269721553785044e-05,
+      "loss": 4.0664,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926135198507825e-05,
+      "loss": 4.063,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925296603756773e-05,
+      "loss": 4.0647,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924458009005721e-05,
+      "loss": 4.0753,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923619414254669e-05,
+      "loss": 4.0655,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922780819503617e-05,
+      "loss": 4.061,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921942224752565e-05,
+      "loss": 4.0698,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921103630001514e-05,
+      "loss": 4.078,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920265035250462e-05,
+      "loss": 4.0679,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919428078379783e-05,
+      "loss": 4.0592,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918589483628731e-05,
+      "loss": 4.0618,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917750888877679e-05,
+      "loss": 4.064,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916912294126627e-05,
+      "loss": 4.0746,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916073699375575e-05,
+      "loss": 4.0741,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915235104624523e-05,
+      "loss": 4.0654,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914398147753844e-05,
+      "loss": 4.0642,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913559553002792e-05,
+      "loss": 4.0573,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91272095825174e-05,
+      "loss": 4.0611,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911882363500688e-05,
+      "loss": 4.0646,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911043768749636e-05,
+      "loss": 4.0517,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910205173998584e-05,
+      "loss": 4.0733,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909366579247532e-05,
+      "loss": 4.0597,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90852798449648e-05,
+      "loss": 4.0607,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907689389745428e-05,
+      "loss": 4.0594,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906852432874749e-05,
+      "loss": 4.0552,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060138381236966e-05,
+      "loss": 4.073,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9051752433726446e-05,
+      "loss": 4.0577,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9043366486215926e-05,
+      "loss": 4.082,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903499691750914e-05,
+      "loss": 4.0719,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026610969998615e-05,
+      "loss": 4.0642,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901824140129183e-05,
+      "loss": 4.0661,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900985545378131e-05,
+      "loss": 4.061,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900146950627079e-05,
+      "loss": 4.0622,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899308355876027e-05,
+      "loss": 4.0714,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898469761124975e-05,
+      "loss": 4.0602,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897631166373923e-05,
+      "loss": 4.0656,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896792571622871e-05,
+      "loss": 4.0601,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895953976871819e-05,
+      "loss": 4.0648,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895115382120767e-05,
+      "loss": 4.0623,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894278425250088e-05,
+      "loss": 4.0538,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893441468379409e-05,
+      "loss": 4.0689,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892602873628357e-05,
+      "loss": 4.0514,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891764278877305e-05,
+      "loss": 4.0747,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890925684126253e-05,
+      "loss": 4.0635,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890087089375201e-05,
+      "loss": 4.0658,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892484946241495e-05,
+      "loss": 4.0529,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884098998730975e-05,
+      "loss": 4.0724,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8875729430024184e-05,
+      "loss": 4.0631,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8867343482513664e-05,
+      "loss": 4.0678,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8858957535003144e-05,
+      "loss": 4.0601,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8850571587492624e-05,
+      "loss": 4.0684,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8842185639982104e-05,
+      "loss": 4.0623,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8833799692471584e-05,
+      "loss": 4.0802,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8825413744961064e-05,
+      "loss": 4.0532,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817027797450544e-05,
+      "loss": 4.0705,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880865822874375e-05,
+      "loss": 4.0641,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880027228123323e-05,
+      "loss": 4.0673,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879190271252645e-05,
+      "loss": 4.0572,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878351676501593e-05,
+      "loss": 4.0668,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877513081750541e-05,
+      "loss": 4.0633,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876674486999489e-05,
+      "loss": 4.0552,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875835892248437e-05,
+      "loss": 4.0654,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.089563846588135,
+      "eval_runtime": 294.4954,
+      "eval_samples_per_second": 1295.745,
+      "eval_steps_per_second": 40.493,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874997297497385e-05,
+      "loss": 4.0553,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874158702746333e-05,
+      "loss": 4.0517,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87332010799528e-05,
+      "loss": 4.0613,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872481513244228e-05,
+      "loss": 4.0734,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871642918493176e-05,
+      "loss": 4.067,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870804323742124e-05,
+      "loss": 4.0552,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869965728991072e-05,
+      "loss": 4.072,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86912713424002e-05,
+      "loss": 4.0502,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868288539488968e-05,
+      "loss": 4.0612,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867449944737916e-05,
+      "loss": 4.0722,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866611349986864e-05,
+      "loss": 4.0597,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865772755235813e-05,
+      "loss": 4.0639,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864935798365134e-05,
+      "loss": 4.0663,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864097203614082e-05,
+      "loss": 4.0536,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86325860886303e-05,
+      "loss": 4.0507,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.862420014111978e-05,
+      "loss": 4.0487,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861581419360926e-05,
+      "loss": 4.0562,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.860742824609874e-05,
+      "loss": 4.0577,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859904229858822e-05,
+      "loss": 4.0545,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859065635107769e-05,
+      "loss": 4.0669,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858227040356717e-05,
+      "loss": 4.0705,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857388445605665e-05,
+      "loss": 4.0693,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8565514887349866e-05,
+      "loss": 4.0618,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8557128939839346e-05,
+      "loss": 4.0563,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8548742992328826e-05,
+      "loss": 4.0627,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8540357044818306e-05,
+      "loss": 4.056,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8531971097307786e-05,
+      "loss": 4.0556,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8523585149797266e-05,
+      "loss": 4.0523,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8515199202286746e-05,
+      "loss": 4.0542,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8506813254776226e-05,
+      "loss": 4.0521,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8498443686069435e-05,
+      "loss": 4.0508,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8490057738558915e-05,
+      "loss": 4.0603,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8481671791048395e-05,
+      "loss": 4.0682,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8473285843537875e-05,
+      "loss": 4.0567,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8464916274831084e-05,
+      "loss": 4.0562,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84565467061243e-05,
+      "loss": 4.0558,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844816075861378e-05,
+      "loss": 4.0626,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843977481110326e-05,
+      "loss": 4.0474,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843138886359274e-05,
+      "loss": 4.055,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842300291608222e-05,
+      "loss": 4.0421,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84146169685717e-05,
+      "loss": 4.0558,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840623102106118e-05,
+      "loss": 4.0556,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839784507355066e-05,
+      "loss": 4.0501,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83894918836476e-05,
+      "loss": 4.0553,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838110593613708e-05,
+      "loss": 4.0617,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837271998862656e-05,
+      "loss": 4.0577,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836433404111604e-05,
+      "loss": 4.0611,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835594809360552e-05,
+      "loss": 4.0531,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8347562146095004e-05,
+      "loss": 4.025,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8339176198584484e-05,
+      "loss": 4.0645,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8330790251073964e-05,
+      "loss": 4.0557,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832242068236717e-05,
+      "loss": 4.0596,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831403473485665e-05,
+      "loss": 4.0551,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830564878734613e-05,
+      "loss": 4.0429,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829726283983561e-05,
+      "loss": 4.0391,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828889327112882e-05,
+      "loss": 4.0459,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82805073236183e-05,
+      "loss": 4.0455,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827212137610778e-05,
+      "loss": 4.0398,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826373542859726e-05,
+      "loss": 4.064,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825534948108674e-05,
+      "loss": 4.0475,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824696353357622e-05,
+      "loss": 4.0306,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82385775860657e-05,
+      "loss": 4.0636,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823019163855518e-05,
+      "loss": 4.0353,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82218220698484e-05,
+      "loss": 4.0366,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821343612233788e-05,
+      "loss": 4.0528,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820505017482735e-05,
+      "loss": 4.0494,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819666422731683e-05,
+      "loss": 4.0443,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818827827980631e-05,
+      "loss": 4.042,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817990871109952e-05,
+      "loss": 4.0375,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171522763589e-05,
+      "loss": 4.0509,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8163153194882216e-05,
+      "loss": 4.0473,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154767247371695e-05,
+      "loss": 4.0501,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8146381299861175e-05,
+      "loss": 4.0479,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8137995352350655e-05,
+      "loss": 4.0505,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8129609404840135e-05,
+      "loss": 4.0639,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8121223457329615e-05,
+      "loss": 4.0351,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8112853888622824e-05,
+      "loss": 4.0492,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8104467941112304e-05,
+      "loss": 4.058,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8096081993601784e-05,
+      "loss": 4.0309,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8087696046091264e-05,
+      "loss": 4.0457,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8079310098580744e-05,
+      "loss": 4.0516,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8070924151070224e-05,
+      "loss": 4.0564,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806255458236343e-05,
+      "loss": 4.0545,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805416863485291e-05,
+      "loss": 4.0322,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804578268734239e-05,
+      "loss": 4.0277,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803739673983187e-05,
+      "loss": 4.0485,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802901079232136e-05,
+      "loss": 4.0494,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802064122361457e-05,
+      "loss": 4.0402,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801227165490778e-05,
+      "loss": 4.0442,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800388570739726e-05,
+      "loss": 4.0392,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799549975988674e-05,
+      "loss": 4.0501,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798711381237622e-05,
+      "loss": 4.044,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797874424366943e-05,
+      "loss": 4.0385,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797035829615891e-05,
+      "loss": 4.0423,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796197234864839e-05,
+      "loss": 4.0536,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795358640113787e-05,
+      "loss": 4.0481,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794520045362735e-05,
+      "loss": 4.0363,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793681450611683e-05,
+      "loss": 4.0369,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7928428558606314e-05,
+      "loss": 4.0406,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920042611095793e-05,
+      "loss": 4.0504,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911656663585273e-05,
+      "loss": 4.0487,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7903270716074753e-05,
+      "loss": 4.0414,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789488476856423e-05,
+      "loss": 4.0428,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788649882105371e-05,
+      "loss": 4.0335,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7878112873543186e-05,
+      "loss": 4.0418,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78697433048364e-05,
+      "loss": 4.0353,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786135735732588e-05,
+      "loss": 4.0298,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7852971409815355e-05,
+      "loss": 4.0494,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784460184110857e-05,
+      "loss": 4.0346,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783621589359805e-05,
+      "loss": 4.0388,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782782994608753e-05,
+      "loss": 4.0344,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781944399857701e-05,
+      "loss": 4.0319,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781107442987023e-05,
+      "loss": 4.0486,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780268848235971e-05,
+      "loss": 4.036,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779430253484919e-05,
+      "loss": 4.0574,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778591658733866e-05,
+      "loss": 4.0489,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777753063982814e-05,
+      "loss": 4.0426,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776914469231762e-05,
+      "loss": 4.0421,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776077512361083e-05,
+      "loss": 4.0387,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775238917610031e-05,
+      "loss": 4.036,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774400322858979e-05,
+      "loss": 4.0492,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773561728107927e-05,
+      "loss": 4.0424,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772723133356875e-05,
+      "loss": 4.0392,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7718845386058236e-05,
+      "loss": 4.0346,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7710459438547716e-05,
+      "loss": 4.0403,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7702073491037196e-05,
+      "loss": 4.0379,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7693703922330405e-05,
+      "loss": 4.0309,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7685334353623614e-05,
+      "loss": 4.0466,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7676948406113094e-05,
+      "loss": 4.0235,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7668562458602574e-05,
+      "loss": 4.0584,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7660176511092054e-05,
+      "loss": 4.0385,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7651790563581534e-05,
+      "loss": 4.0426,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7643404616071014e-05,
+      "loss": 4.0288,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7635018668560494e-05,
+      "loss": 4.05,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7626632721049974e-05,
+      "loss": 4.0407,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761826315234319e-05,
+      "loss": 4.0438,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760987720483267e-05,
+      "loss": 4.0386,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760150763612588e-05,
+      "loss": 4.0407,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759312168861536e-05,
+      "loss": 4.0397,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758473574110484e-05,
+      "loss": 4.0587,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757634979359432e-05,
+      "loss": 4.0301,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75679638460838e-05,
+      "loss": 4.0473,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755957789857328e-05,
+      "loss": 4.0451,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755119195106276e-05,
+      "loss": 4.0419,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754280600355224e-05,
+      "loss": 4.0325,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753443643484545e-05,
+      "loss": 4.0438,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752605048733493e-05,
+      "loss": 4.0394,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751768091862814e-05,
+      "loss": 4.0356,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.750929497111762e-05,
+      "loss": 4.0421,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.073797225952148,
+      "eval_runtime": 292.9055,
+      "eval_samples_per_second": 1302.779,
+      "eval_steps_per_second": 40.713,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.75009090236071e-05,
+      "loss": 4.031,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749252307609658e-05,
+      "loss": 4.0271,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748413712858606e-05,
+      "loss": 4.0404,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.747575118107554e-05,
+      "loss": 4.0486,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7467365233565016e-05,
+      "loss": 4.0453,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7458979286054496e-05,
+      "loss": 4.0343,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7450593338543976e-05,
+      "loss": 4.0434,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744224014864092e-05,
+      "loss": 4.0326,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74338542011304e-05,
+      "loss": 4.0341,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742546825361988e-05,
+      "loss": 4.0489,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.741708230610936e-05,
+      "loss": 4.0397,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740869635859884e-05,
+      "loss": 4.0436,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740032678989206e-05,
+      "loss": 4.0456,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739194084238154e-05,
+      "loss": 4.0297,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7383554894871017e-05,
+      "loss": 4.0272,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737516894736049e-05,
+      "loss": 4.0215,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736678299984997e-05,
+      "loss": 4.0348,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735839705233945e-05,
+      "loss": 4.0386,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735001110482893e-05,
+      "loss": 4.0286,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734162515731841e-05,
+      "loss": 4.0434,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733325558861162e-05,
+      "loss": 4.049,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.73248696411011e-05,
+      "loss": 4.0467,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731648369359058e-05,
+      "loss": 4.0415,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730809774608006e-05,
+      "loss": 4.036,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7299711798569545e-05,
+      "loss": 4.0404,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7291325851059025e-05,
+      "loss": 4.0369,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7282956282352234e-05,
+      "loss": 4.0307,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7274570334841714e-05,
+      "loss": 4.0278,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7266184387331194e-05,
+      "loss": 4.035,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7257798439820674e-05,
+      "loss": 4.0309,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7249412492310154e-05,
+      "loss": 4.0319,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724105930240709e-05,
+      "loss": 4.0374,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.723267335489657e-05,
+      "loss": 4.0412,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722428740738605e-05,
+      "loss": 4.0419,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721590145987553e-05,
+      "loss": 4.0275,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720753189116875e-05,
+      "loss": 4.0388,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719914594365823e-05,
+      "loss": 4.0412,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719075999614771e-05,
+      "loss": 4.0213,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718237404863719e-05,
+      "loss": 4.0344,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717398810112667e-05,
+      "loss": 4.0222,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716560215361615e-05,
+      "loss": 4.0337,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715721620610563e-05,
+      "loss": 4.0321,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714883025859511e-05,
+      "loss": 4.0278,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714044431108459e-05,
+      "loss": 4.0332,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713205836357407e-05,
+      "loss": 4.041,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712367241606355e-05,
+      "loss": 4.0375,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711528646855303e-05,
+      "loss": 4.0409,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710691689984624e-05,
+      "loss": 4.0299,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709853095233572e-05,
+      "loss": 4.0064,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70901450048252e-05,
+      "loss": 4.0418,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7081759057314677e-05,
+      "loss": 4.0338,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707338948860789e-05,
+      "loss": 4.0379,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706500354109737e-05,
+      "loss": 4.0307,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705663397239058e-05,
+      "loss": 4.0242,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704824802488006e-05,
+      "loss": 4.0154,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703987845617327e-05,
+      "loss": 4.0235,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703149250866275e-05,
+      "loss": 4.0271,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702310656115223e-05,
+      "loss": 4.0136,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701472061364171e-05,
+      "loss": 4.0432,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700635104493492e-05,
+      "loss": 4.0284,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6997965097424406e-05,
+      "loss": 4.011,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6989579149913886e-05,
+      "loss": 4.038,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6981193202403366e-05,
+      "loss": 4.0158,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6972807254892846e-05,
+      "loss": 4.0166,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6964421307382326e-05,
+      "loss": 4.0315,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69560353598718e-05,
+      "loss": 4.0275,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694764941236128e-05,
+      "loss": 4.0211,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693926346485076e-05,
+      "loss": 4.0194,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693087751734024e-05,
+      "loss": 4.0173,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692249156982972e-05,
+      "loss": 4.0252,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69141056223192e-05,
+      "loss": 4.0346,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690571967480868e-05,
+      "loss": 4.025,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689733372729816e-05,
+      "loss": 4.0271,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688894777978764e-05,
+      "loss": 4.0247,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688056183227712e-05,
+      "loss": 4.0438,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6872192263570335e-05,
+      "loss": 4.0187,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6863806316059815e-05,
+      "loss": 4.0294,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6855420368549295e-05,
+      "loss": 4.0302,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6847034421038775e-05,
+      "loss": 4.0159,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683868123113571e-05,
+      "loss": 4.0239,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683029528362519e-05,
+      "loss": 4.031,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682190933611467e-05,
+      "loss": 4.0351,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681352338860415e-05,
+      "loss": 4.0382,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680513744109363e-05,
+      "loss": 4.0078,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679675149358311e-05,
+      "loss": 4.0115,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678836554607259e-05,
+      "loss": 4.0253,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677997959856207e-05,
+      "loss": 4.0274,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677159365105155e-05,
+      "loss": 4.0192,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676322408234477e-05,
+      "loss": 4.0247,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675483813483425e-05,
+      "loss": 4.0218,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674645218732373e-05,
+      "loss": 4.0301,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673808261861694e-05,
+      "loss": 4.0217,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672969667110642e-05,
+      "loss": 4.0156,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67213107235959e-05,
+      "loss": 4.024,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671292477608538e-05,
+      "loss": 4.032,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6704555207378586e-05,
+      "loss": 4.0282,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6696169259868066e-05,
+      "loss": 4.0134,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6687799691161275e-05,
+      "loss": 4.0144,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667941374365076e-05,
+      "loss": 4.022,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667102779614024e-05,
+      "loss": 4.0284,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666264184862972e-05,
+      "loss": 4.0301,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66542559011192e-05,
+      "loss": 4.0185,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664586995360868e-05,
+      "loss": 4.0209,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663748400609816e-05,
+      "loss": 4.0114,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6629098058587635e-05,
+      "loss": 4.0184,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6620712111077115e-05,
+      "loss": 4.0152,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6612326163566595e-05,
+      "loss": 4.0135,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6603940216056075e-05,
+      "loss": 4.0256,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6595554268545555e-05,
+      "loss": 4.0146,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6587184699838764e-05,
+      "loss": 4.0177,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6578798752328244e-05,
+      "loss": 4.0128,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657042918362146e-05,
+      "loss": 4.0098,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.656204323611094e-05,
+      "loss": 4.0252,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.655365728860042e-05,
+      "loss": 4.0232,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65452713410899e-05,
+      "loss": 4.0295,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653688539357938e-05,
+      "loss": 4.0267,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652851582487259e-05,
+      "loss": 4.0247,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652012987736207e-05,
+      "loss": 4.0227,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651174392985155e-05,
+      "loss": 4.0174,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650335798234103e-05,
+      "loss": 4.0185,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649497203483051e-05,
+      "loss": 4.0273,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648658608731999e-05,
+      "loss": 4.0199,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64782165186132e-05,
+      "loss": 4.0239,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6469830571102684e-05,
+      "loss": 4.0101,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6461444623592164e-05,
+      "loss": 4.0235,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6453058676081644e-05,
+      "loss": 4.0148,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6444672728571124e-05,
+      "loss": 4.013,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643630315986433e-05,
+      "loss": 4.0214,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642791721235381e-05,
+      "loss": 4.0034,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641953126484329e-05,
+      "loss": 4.0396,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641114531733277e-05,
+      "loss": 4.0171,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640275936982225e-05,
+      "loss": 4.0258,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639437342231173e-05,
+      "loss": 4.0084,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638600385360494e-05,
+      "loss": 4.0261,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637761790609442e-05,
+      "loss": 4.0204,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63692319585839e-05,
+      "loss": 4.0262,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636084601107338e-05,
+      "loss": 4.0186,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635246006356286e-05,
+      "loss": 4.0174,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634407411605235e-05,
+      "loss": 4.0221,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.633568816854182e-05,
+      "loss": 4.0417,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632731859983504e-05,
+      "loss": 4.0055,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631893265232452e-05,
+      "loss": 4.0317,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6310546704814e-05,
+      "loss": 4.028,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630216075730347e-05,
+      "loss": 4.0191,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629377480979295e-05,
+      "loss": 4.0139,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628538886228243e-05,
+      "loss": 4.0229,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627700291477191e-05,
+      "loss": 4.0197,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626861696726139e-05,
+      "loss": 4.0181,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.62602473985546e-05,
+      "loss": 4.0183,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.060302734375,
+      "eval_runtime": 297.2497,
+      "eval_samples_per_second": 1283.739,
+      "eval_steps_per_second": 40.118,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6251861451044087e-05,
+      "loss": 4.0122,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6243475503533566e-05,
+      "loss": 4.0089,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6235089556023046e-05,
+      "loss": 4.0175,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6226703608512526e-05,
+      "loss": 4.0283,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6218317661002006e-05,
+      "loss": 4.0276,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6209931713491486e-05,
+      "loss": 4.0148,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6201545765980966e-05,
+      "loss": 4.0254,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6193159818470446e-05,
+      "loss": 4.015,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6184773870959926e-05,
+      "loss": 4.0116,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6176387923449406e-05,
+      "loss": 4.025,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6168001975938886e-05,
+      "loss": 4.0262,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615961602842836e-05,
+      "loss": 4.024,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6151246459721575e-05,
+      "loss": 4.0202,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6142860512211055e-05,
+      "loss": 4.0099,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6134474564700535e-05,
+      "loss": 4.0131,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6126088617190015e-05,
+      "loss": 3.9987,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6117702669679495e-05,
+      "loss": 4.0191,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6109316722168975e-05,
+      "loss": 4.0176,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6100930774658455e-05,
+      "loss": 4.0083,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6092544827147935e-05,
+      "loss": 4.021,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6084158879637415e-05,
+      "loss": 4.0302,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6075772932126895e-05,
+      "loss": 4.0241,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6067386984616375e-05,
+      "loss": 4.0234,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6059001037105855e-05,
+      "loss": 4.0182,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6050631468399064e-05,
+      "loss": 4.0203,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6042245520888544e-05,
+      "loss": 4.0188,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6033859573378024e-05,
+      "loss": 4.0111,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6025473625867504e-05,
+      "loss": 4.0052,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6017087678356984e-05,
+      "loss": 4.0166,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6008701730846464e-05,
+      "loss": 4.0109,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600031578333595e-05,
+      "loss": 4.0153,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5991929835825423e-05,
+      "loss": 4.0173,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5983543888314903e-05,
+      "loss": 4.0183,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597515794080438e-05,
+      "loss": 4.0262,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596680475090133e-05,
+      "loss": 4.0075,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595843518219454e-05,
+      "loss": 4.0183,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595004923468402e-05,
+      "loss": 4.0215,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59416632871735e-05,
+      "loss": 4.0001,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593327733966298e-05,
+      "loss": 4.0151,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592489139215246e-05,
+      "loss": 4.005,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591650544464194e-05,
+      "loss": 4.0084,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590811949713142e-05,
+      "loss": 4.0142,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58997335496209e-05,
+      "loss": 4.0095,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589136398091411e-05,
+      "loss": 4.0143,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588297803340359e-05,
+      "loss": 4.0198,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587459208589307e-05,
+      "loss": 4.02,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5866206138382546e-05,
+      "loss": 4.0201,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5857820190872026e-05,
+      "loss": 4.0135,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5849434243361506e-05,
+      "loss": 3.9879,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5841048295850986e-05,
+      "loss": 4.0217,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5832662348340466e-05,
+      "loss": 4.0143,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5824292779633675e-05,
+      "loss": 4.0224,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581592321092689e-05,
+      "loss": 4.0131,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580753726341637e-05,
+      "loss": 4.0017,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579915131590585e-05,
+      "loss": 4.0042,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579076536839533e-05,
+      "loss": 3.9984,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578239579968855e-05,
+      "loss": 4.0115,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.577400985217802e-05,
+      "loss": 3.9912,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57656239046675e-05,
+      "loss": 4.028,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5757254335960716e-05,
+      "loss": 4.0094,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574886838845019e-05,
+      "loss": 3.9908,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574048244093967e-05,
+      "loss": 4.0197,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573209649342915e-05,
+      "loss": 3.998,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572371054591863e-05,
+      "loss": 3.9997,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571532459840811e-05,
+      "loss": 4.008,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5706938650897595e-05,
+      "loss": 4.0062,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5698552703387075e-05,
+      "loss": 4.0021,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5690166755876555e-05,
+      "loss": 4.0055,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5681813565973494e-05,
+      "loss": 3.9962,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5673427618462974e-05,
+      "loss": 4.0076,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5665041670952453e-05,
+      "loss": 4.0151,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5656655723441933e-05,
+      "loss": 4.0056,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564826977593141e-05,
+      "loss": 4.0127,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563988382842089e-05,
+      "loss": 4.0026,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563149788091037e-05,
+      "loss": 4.0291,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562311193339985e-05,
+      "loss": 3.9957,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.561472598588933e-05,
+      "loss": 4.0136,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560634003837881e-05,
+      "loss": 4.004,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559795409086829e-05,
+      "loss": 4.0051,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558958452216151e-05,
+      "loss": 4.0039,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558119857465099e-05,
+      "loss": 4.0117,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557281262714047e-05,
+      "loss": 4.0169,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556442667962995e-05,
+      "loss": 4.0174,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555604073211943e-05,
+      "loss": 3.9904,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554765478460891e-05,
+      "loss": 3.9929,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553928521590212e-05,
+      "loss": 4.0052,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55308992683916e-05,
+      "loss": 4.0124,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552251332088108e-05,
+      "loss": 4.0017,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.551412737337055e-05,
+      "loss": 4.0039,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550575780466377e-05,
+      "loss": 4.004,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549737185715325e-05,
+      "loss": 4.0067,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5488985909642733e-05,
+      "loss": 4.008,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548059996213221e-05,
+      "loss": 4.0029,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547221401462169e-05,
+      "loss": 4.0065,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5463828067111167e-05,
+      "loss": 4.0127,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5455442119600647e-05,
+      "loss": 4.0132,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5447056172090126e-05,
+      "loss": 3.9957,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5438686603383336e-05,
+      "loss": 3.9956,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5430300655872816e-05,
+      "loss": 4.0041,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5421914708362296e-05,
+      "loss": 4.0087,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5413528760851775e-05,
+      "loss": 4.0136,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5405142813341255e-05,
+      "loss": 4.001,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.539677324463447e-05,
+      "loss": 4.0043,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538838729712395e-05,
+      "loss": 3.9959,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538000134961343e-05,
+      "loss": 4.0014,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537163178090664e-05,
+      "loss": 4.0007,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.536324583339612e-05,
+      "loss": 3.9962,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53548598858856e-05,
+      "loss": 4.0051,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534647393837508e-05,
+      "loss": 4.0002,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533808799086456e-05,
+      "loss": 3.9991,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532970204335404e-05,
+      "loss": 3.9987,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532131609584352e-05,
+      "loss": 3.9918,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531294652713673e-05,
+      "loss": 4.0092,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530456057962621e-05,
+      "loss": 4.0046,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529617463211569e-05,
+      "loss": 4.0118,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528778868460517e-05,
+      "loss": 4.0061,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527940273709465e-05,
+      "loss": 4.0095,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5271016789584136e-05,
+      "loss": 4.0054,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5262630842073616e-05,
+      "loss": 3.9962,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5254261273366825e-05,
+      "loss": 4.006,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5245875325856305e-05,
+      "loss": 4.0055,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5237489378345785e-05,
+      "loss": 4.0009,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5229119809638994e-05,
+      "loss": 4.0079,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5220733862128474e-05,
+      "loss": 3.9987,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5212347914617954e-05,
+      "loss": 4.0027,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5203961967107434e-05,
+      "loss": 3.9966,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5195576019596914e-05,
+      "loss": 3.9981,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518720645089012e-05,
+      "loss": 4.0025,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51788205033796e-05,
+      "loss": 3.9915,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517043455586909e-05,
+      "loss": 4.0194,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516204860835857e-05,
+      "loss": 4.0012,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515366266084804e-05,
+      "loss": 4.0065,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.514527671333752e-05,
+      "loss": 3.9932,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5136890765827e-05,
+      "loss": 4.0078,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512852119712021e-05,
+      "loss": 4.0044,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512013524960969e-05,
+      "loss": 4.0078,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511174930209917e-05,
+      "loss": 4.0009,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510336335458865e-05,
+      "loss": 3.9999,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509497740707813e-05,
+      "loss": 4.0022,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508659145956761e-05,
+      "loss": 4.0236,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507820551205709e-05,
+      "loss": 3.9934,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506983594335031e-05,
+      "loss": 4.0132,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506144999583979e-05,
+      "loss": 4.0115,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505306404832927e-05,
+      "loss": 4.0003,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5044694479622476e-05,
+      "loss": 4.0039,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5036308532111956e-05,
+      "loss": 3.9998,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5027922584601436e-05,
+      "loss": 4.0014,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5019536637090916e-05,
+      "loss": 4.001,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5011150689580396e-05,
+      "loss": 4.0005,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.049688816070557,
+      "eval_runtime": 298.266,
+      "eval_samples_per_second": 1279.365,
+      "eval_steps_per_second": 39.981,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5002764742069876e-05,
+      "loss": 3.9998,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4994378794559356e-05,
+      "loss": 3.9865,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4985992847048836e-05,
+      "loss": 3.9976,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4977606899538316e-05,
+      "loss": 4.0147,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4969220952027796e-05,
+      "loss": 4.0047,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4960835004517276e-05,
+      "loss": 4.0002,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4952449057006756e-05,
+      "loss": 4.013,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.494407948829997e-05,
+      "loss": 3.9951,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.493569354078945e-05,
+      "loss": 3.9918,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492730759327893e-05,
+      "loss": 4.0107,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4918921645768405e-05,
+      "loss": 4.0075,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491055207706162e-05,
+      "loss": 4.0033,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.490218250835483e-05,
+      "loss": 4.0108,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.489379656084431e-05,
+      "loss": 3.9906,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488541061333379e-05,
+      "loss": 3.9938,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487702466582327e-05,
+      "loss": 3.9846,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486863871831275e-05,
+      "loss": 4.0022,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486025277080223e-05,
+      "loss": 4.0019,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4851883202095445e-05,
+      "loss": 3.9938,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4843497254584925e-05,
+      "loss": 3.9976,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4835111307074405e-05,
+      "loss": 4.0205,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482672535956388e-05,
+      "loss": 4.0022,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481833941205336e-05,
+      "loss": 4.0083,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.480995346454284e-05,
+      "loss": 4.0018,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.480158389583605e-05,
+      "loss": 4.0059,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.479319794832553e-05,
+      "loss": 3.999,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.478481200081501e-05,
+      "loss": 3.9972,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.477642605330449e-05,
+      "loss": 3.9877,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476804010579397e-05,
+      "loss": 4.0009,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.475965415828345e-05,
+      "loss": 3.9903,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.475126821077293e-05,
+      "loss": 3.9996,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.474289864206614e-05,
+      "loss": 3.9975,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.473451269455562e-05,
+      "loss": 4.0068,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.47261267470451e-05,
+      "loss": 4.0045,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471774079953458e-05,
+      "loss": 3.9947,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470937123082779e-05,
+      "loss": 4.0029,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470098528331727e-05,
+      "loss": 4.0055,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.469259933580675e-05,
+      "loss": 3.9812,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468421338829623e-05,
+      "loss": 3.999,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467582744078571e-05,
+      "loss": 3.9899,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466744149327519e-05,
+      "loss": 3.9899,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465905554576467e-05,
+      "loss": 4.0015,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465066959825415e-05,
+      "loss": 3.9952,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464228365074363e-05,
+      "loss": 3.9978,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463391408203685e-05,
+      "loss": 4.0001,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462552813452633e-05,
+      "loss": 4.0021,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461714218701581e-05,
+      "loss": 4.0038,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460875623950529e-05,
+      "loss": 3.9976,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460037029199476e-05,
+      "loss": 3.9707,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.459198434448424e-05,
+      "loss": 4.0022,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.458359839697372e-05,
+      "loss": 3.999,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45752124494632e-05,
+      "loss": 4.0056,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.456682650195268e-05,
+      "loss": 3.9973,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455845693324589e-05,
+      "loss": 3.9883,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455007098573537e-05,
+      "loss": 3.9818,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4541701417028585e-05,
+      "loss": 3.9849,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4533315469518065e-05,
+      "loss": 3.991,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4524929522007545e-05,
+      "loss": 3.9763,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4516543574497025e-05,
+      "loss": 4.0125,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4508157626986505e-05,
+      "loss": 3.9966,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4499771679475985e-05,
+      "loss": 3.9748,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4491385731965465e-05,
+      "loss": 4.0006,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4482999784454945e-05,
+      "loss": 3.9822,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4474630215748154e-05,
+      "loss": 3.9831,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4466244268237634e-05,
+      "loss": 3.992,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4457858320727114e-05,
+      "loss": 3.9939,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4449472373216594e-05,
+      "loss": 3.9804,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44411028045098e-05,
+      "loss": 3.9941,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443271685699929e-05,
+      "loss": 3.9782,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442433090948877e-05,
+      "loss": 3.9883,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.441594496197825e-05,
+      "loss": 3.9979,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440755901446773e-05,
+      "loss": 3.9915,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439917306695721e-05,
+      "loss": 3.9916,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439078711944669e-05,
+      "loss": 3.9882,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.438240117193617e-05,
+      "loss": 4.013,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.437403160322938e-05,
+      "loss": 3.9774,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.436564565571886e-05,
+      "loss": 3.998,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435725970820834e-05,
+      "loss": 3.9877,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434889013950155e-05,
+      "loss": 3.9884,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434050419199103e-05,
+      "loss": 3.9909,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.433211824448051e-05,
+      "loss": 3.9898,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.432373229696999e-05,
+      "loss": 4.0046,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.431534634945947e-05,
+      "loss": 4.0015,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4306960401948954e-05,
+      "loss": 3.9764,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429857445443843e-05,
+      "loss": 3.9767,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429018850692791e-05,
+      "loss": 3.9843,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.428181893822112e-05,
+      "loss": 3.9972,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4273432990710596e-05,
+      "loss": 3.9868,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4265047043200076e-05,
+      "loss": 3.9844,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4256661095689556e-05,
+      "loss": 3.9938,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4248275148179036e-05,
+      "loss": 3.9878,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.423992195827598e-05,
+      "loss": 3.9926,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.42315523895692e-05,
+      "loss": 3.9822,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422316644205868e-05,
+      "loss": 3.9923,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.421478049454816e-05,
+      "loss": 3.9944,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.420639454703764e-05,
+      "loss": 3.9976,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.419800859952712e-05,
+      "loss": 3.9791,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41896226520166e-05,
+      "loss": 3.98,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418123670450607e-05,
+      "loss": 3.9908,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.417285075699555e-05,
+      "loss": 3.9898,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.416446480948503e-05,
+      "loss": 4.0015,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.415607886197451e-05,
+      "loss": 3.9819,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414770929326772e-05,
+      "loss": 3.9881,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41393233457572e-05,
+      "loss": 3.9828,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413093739824668e-05,
+      "loss": 3.9801,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.412255145073616e-05,
+      "loss": 3.9864,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4114181882029375e-05,
+      "loss": 3.9849,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4105795934518855e-05,
+      "loss": 3.9883,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4097409987008335e-05,
+      "loss": 3.9822,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4089024039497815e-05,
+      "loss": 3.9834,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4080638091987294e-05,
+      "loss": 3.9832,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4072252144476774e-05,
+      "loss": 3.973,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4063866196966254e-05,
+      "loss": 3.9912,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4055480249455734e-05,
+      "loss": 3.9915,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4047110680748943e-05,
+      "loss": 3.9919,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4038724733238423e-05,
+      "loss": 3.9964,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40303387857279e-05,
+      "loss": 3.9939,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.402195283821738e-05,
+      "loss": 3.9876,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.401356689070686e-05,
+      "loss": 3.981,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.400519732200008e-05,
+      "loss": 3.9923,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.399682775329329e-05,
+      "loss": 3.9858,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398844180578277e-05,
+      "loss": 3.9859,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398005585827225e-05,
+      "loss": 3.9914,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.397166991076173e-05,
+      "loss": 3.9807,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396328396325121e-05,
+      "loss": 3.9888,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.395489801574069e-05,
+      "loss": 3.9781,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394651206823017e-05,
+      "loss": 3.9829,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393812612071965e-05,
+      "loss": 3.9875,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392975655201286e-05,
+      "loss": 3.9791,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392137060450234e-05,
+      "loss": 4.0004,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391298465699182e-05,
+      "loss": 3.9881,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390461508828503e-05,
+      "loss": 3.9899,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389622914077451e-05,
+      "loss": 3.9789,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388784319326399e-05,
+      "loss": 3.9866,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387945724575347e-05,
+      "loss": 3.9915,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387107129824295e-05,
+      "loss": 3.9928,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.386268535073243e-05,
+      "loss": 3.9861,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3854299403221906e-05,
+      "loss": 3.9819,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3845913455711386e-05,
+      "loss": 3.989,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3837527508200866e-05,
+      "loss": 4.005,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382915793949408e-05,
+      "loss": 3.9838,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3820771991983555e-05,
+      "loss": 3.9935,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3812386044473035e-05,
+      "loss": 3.9951,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.380400009696252e-05,
+      "loss": 3.9852,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.379563052825573e-05,
+      "loss": 3.9897,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.378724458074521e-05,
+      "loss": 3.9839,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377885863323469e-05,
+      "loss": 3.9872,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377047268572417e-05,
+      "loss": 3.9899,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.376208673821365e-05,
+      "loss": 3.9827,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.040598392486572,
+      "eval_runtime": 297.2975,
+      "eval_samples_per_second": 1283.532,
+      "eval_steps_per_second": 40.111,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375370079070313e-05,
+      "loss": 4.0038,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374531484319261e-05,
+      "loss": 3.9684,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373692889568209e-05,
+      "loss": 3.9842,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372854294817157e-05,
+      "loss": 4.0023,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372015700066105e-05,
+      "loss": 3.9894,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371177105315053e-05,
+      "loss": 3.9843,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370338510564001e-05,
+      "loss": 3.9918,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369499915812949e-05,
+      "loss": 3.984,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368661321061897e-05,
+      "loss": 3.9767,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367822726310845e-05,
+      "loss": 3.9979,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366984131559793e-05,
+      "loss": 3.9875,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366145536808741e-05,
+      "loss": 3.9918,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365308579938062e-05,
+      "loss": 3.9962,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36446998518701e-05,
+      "loss": 3.9756,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3636330283163315e-05,
+      "loss": 3.9788,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3627944335652795e-05,
+      "loss": 3.969,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361955838814227e-05,
+      "loss": 3.9857,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361117244063175e-05,
+      "loss": 3.9852,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.360278649312123e-05,
+      "loss": 3.9793,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.359440054561071e-05,
+      "loss": 3.9802,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.358601459810019e-05,
+      "loss": 4.0059,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357762865058967e-05,
+      "loss": 3.9872,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3569242703079154e-05,
+      "loss": 3.9905,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3560856755568634e-05,
+      "loss": 3.9884,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3552470808058114e-05,
+      "loss": 3.9911,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3544084860547594e-05,
+      "loss": 3.9819,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3535698913037074e-05,
+      "loss": 3.9808,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352732934433028e-05,
+      "loss": 3.9776,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351895977562349e-05,
+      "loss": 3.98,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351057382811297e-05,
+      "loss": 3.9783,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350218788060245e-05,
+      "loss": 3.9839,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349380193309193e-05,
+      "loss": 3.9791,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348541598558141e-05,
+      "loss": 3.9943,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347703003807089e-05,
+      "loss": 3.9859,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346864409056037e-05,
+      "loss": 3.9816,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346027452185359e-05,
+      "loss": 3.9901,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345188857434307e-05,
+      "loss": 3.9888,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344350262683255e-05,
+      "loss": 3.9637,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343511667932203e-05,
+      "loss": 3.9835,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342673073181151e-05,
+      "loss": 3.9746,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341834478430098e-05,
+      "loss": 3.9747,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340995883679046e-05,
+      "loss": 3.987,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340157288927994e-05,
+      "loss": 3.9826,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.339318694176942e-05,
+      "loss": 3.9787,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338481737306263e-05,
+      "loss": 3.9902,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337643142555211e-05,
+      "loss": 3.9868,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336804547804159e-05,
+      "loss": 3.987,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3359659530531077e-05,
+      "loss": 3.9848,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3351273583020557e-05,
+      "loss": 3.9603,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3342904014313766e-05,
+      "loss": 3.9832,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3334518066803246e-05,
+      "loss": 3.989,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3326132119292726e-05,
+      "loss": 3.9849,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3317746171782206e-05,
+      "loss": 3.9813,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3309360224271685e-05,
+      "loss": 3.9765,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3300990655564895e-05,
+      "loss": 3.9666,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3292621086858104e-05,
+      "loss": 3.9663,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3284235139347584e-05,
+      "loss": 3.9821,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3275849191837064e-05,
+      "loss": 3.9639,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3267463244326544e-05,
+      "loss": 3.9903,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325907729681603e-05,
+      "loss": 3.9815,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325069134930551e-05,
+      "loss": 3.9655,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324230540179499e-05,
+      "loss": 3.9828,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323391945428447e-05,
+      "loss": 3.9669,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322553350677395e-05,
+      "loss": 3.9698,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321714755926343e-05,
+      "loss": 3.9737,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320876161175291e-05,
+      "loss": 3.9824,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320039204304612e-05,
+      "loss": 3.9687,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31920060955356e-05,
+      "loss": 3.9771,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318363652682881e-05,
+      "loss": 3.9624,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317525057931829e-05,
+      "loss": 3.9733,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316686463180777e-05,
+      "loss": 3.9825,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315847868429725e-05,
+      "loss": 3.9791,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315009273678673e-05,
+      "loss": 3.9744,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314170678927621e-05,
+      "loss": 3.9768,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3133320841765695e-05,
+      "loss": 3.9961,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.312493489425517e-05,
+      "loss": 3.9643,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.311654894674465e-05,
+      "loss": 3.9845,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310816299923413e-05,
+      "loss": 3.9737,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309977705172361e-05,
+      "loss": 3.972,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309139110421309e-05,
+      "loss": 3.9733,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30830215355063e-05,
+      "loss": 3.9773,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307463558799578e-05,
+      "loss": 3.9887,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306624964048526e-05,
+      "loss": 3.9898,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305786369297474e-05,
+      "loss": 3.9644,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3049477745464217e-05,
+      "loss": 3.9613,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3041091797953697e-05,
+      "loss": 3.9678,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303272222924691e-05,
+      "loss": 3.9812,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302433628173639e-05,
+      "loss": 3.9716,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30159667130296e-05,
+      "loss": 3.9727,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300758076551908e-05,
+      "loss": 3.9783,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299919481800856e-05,
+      "loss": 3.9718,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299080887049804e-05,
+      "loss": 3.9809,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298242292298752e-05,
+      "loss": 3.9666,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2974036975477e-05,
+      "loss": 3.9782,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296566740677021e-05,
+      "loss": 3.9765,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295728145925969e-05,
+      "loss": 3.9867,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294889551174917e-05,
+      "loss": 3.9635,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294050956423865e-05,
+      "loss": 3.9664,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2932139995531866e-05,
+      "loss": 3.98,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2923754048021346e-05,
+      "loss": 3.974,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2915368100510826e-05,
+      "loss": 3.9848,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2906982153000306e-05,
+      "loss": 3.9697,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2898596205489786e-05,
+      "loss": 3.971,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2890226636782995e-05,
+      "loss": 3.9676,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2881840689272475e-05,
+      "loss": 3.9701,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2873454741761955e-05,
+      "loss": 3.9699,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2865068794251435e-05,
+      "loss": 3.9662,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2856699225544644e-05,
+      "loss": 3.975,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2848313278034124e-05,
+      "loss": 3.9669,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2839927330523604e-05,
+      "loss": 3.9706,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2831541383013084e-05,
+      "loss": 3.969,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282315543550257e-05,
+      "loss": 3.9585,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281478586679578e-05,
+      "loss": 3.9809,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280639991928526e-05,
+      "loss": 3.9755,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279801397177474e-05,
+      "loss": 3.9768,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278962802426422e-05,
+      "loss": 3.9823,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.27812420767537e-05,
+      "loss": 3.9798,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.277285612924318e-05,
+      "loss": 3.9712,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276447018173265e-05,
+      "loss": 3.9654,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275610061302587e-05,
+      "loss": 3.9767,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274771466551535e-05,
+      "loss": 3.9733,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273932871800482e-05,
+      "loss": 3.9688,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273094277049431e-05,
+      "loss": 3.9783,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.272255682298379e-05,
+      "loss": 3.9704,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.271417087547327e-05,
+      "loss": 3.9731,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270578492796275e-05,
+      "loss": 3.9649,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269739898045223e-05,
+      "loss": 3.966,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268902941174544e-05,
+      "loss": 3.9762,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268065984303865e-05,
+      "loss": 3.9659,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2672273895528126e-05,
+      "loss": 3.9783,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2663887948017606e-05,
+      "loss": 3.9826,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265551837931082e-05,
+      "loss": 3.9725,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2647132431800295e-05,
+      "loss": 3.9647,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2638746484289775e-05,
+      "loss": 3.9779,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263036053677926e-05,
+      "loss": 3.9747,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262197458926874e-05,
+      "loss": 3.9767,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261358864175822e-05,
+      "loss": 3.9705,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26052026942477e-05,
+      "loss": 3.9675,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259681674673718e-05,
+      "loss": 3.9808,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258844717803039e-05,
+      "loss": 3.9889,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258006123051987e-05,
+      "loss": 3.9651,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257167528300935e-05,
+      "loss": 3.98,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.256328933549883e-05,
+      "loss": 3.9827,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255491976679204e-05,
+      "loss": 3.9719,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2546566576888985e-05,
+      "loss": 3.9769,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2538180629378465e-05,
+      "loss": 3.9697,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2529794681867945e-05,
+      "loss": 3.9706,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2521408734357425e-05,
+      "loss": 3.9832,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2513022786846905e-05,
+      "loss": 3.9616,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.032841682434082,
+      "eval_runtime": 294.8352,
+      "eval_samples_per_second": 1294.252,
+      "eval_steps_per_second": 40.446,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2504636839336385e-05,
+      "loss": 3.972,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2496250891825865e-05,
+      "loss": 3.96,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2487864944315345e-05,
+      "loss": 3.97,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2479495375608554e-05,
+      "loss": 3.9844,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2471109428098034e-05,
+      "loss": 3.9788,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2462723480587514e-05,
+      "loss": 3.9694,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2454337533076994e-05,
+      "loss": 3.9782,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2445951585566474e-05,
+      "loss": 3.969,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2437565638055953e-05,
+      "loss": 3.9638,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242919606934916e-05,
+      "loss": 3.9824,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242081012183865e-05,
+      "loss": 3.9763,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.241242417432813e-05,
+      "loss": 3.9791,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.240405460562134e-05,
+      "loss": 3.9817,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239566865811082e-05,
+      "loss": 3.9635,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23872827106003e-05,
+      "loss": 3.9627,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237889676308978e-05,
+      "loss": 3.9559,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237051081557926e-05,
+      "loss": 3.9745,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236212486806874e-05,
+      "loss": 3.967,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235373892055822e-05,
+      "loss": 3.9681,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23453529730477e-05,
+      "loss": 3.9719,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233696702553718e-05,
+      "loss": 3.9879,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232859745683039e-05,
+      "loss": 3.9787,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232021150931987e-05,
+      "loss": 3.9722,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.231182556180935e-05,
+      "loss": 3.9753,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2303439614298834e-05,
+      "loss": 3.9791,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2295053666788314e-05,
+      "loss": 3.9705,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228668409808152e-05,
+      "loss": 3.9647,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2278298150571e-05,
+      "loss": 3.9676,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226991220306048e-05,
+      "loss": 3.9621,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2261526255549956e-05,
+      "loss": 3.9669,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2253140308039436e-05,
+      "loss": 3.968,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2244754360528916e-05,
+      "loss": 3.9695,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2236368413018396e-05,
+      "loss": 3.9753,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2227982465507876e-05,
+      "loss": 3.9746,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221964565440856e-05,
+      "loss": 3.9692,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2211259706898037e-05,
+      "loss": 3.9755,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2202873759387517e-05,
+      "loss": 3.977,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2194487811876996e-05,
+      "loss": 3.9515,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2186101864366476e-05,
+      "loss": 3.9761,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2177715916855956e-05,
+      "loss": 3.9541,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216932996934543e-05,
+      "loss": 3.9651,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216094402183491e-05,
+      "loss": 3.9687,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.215255807432439e-05,
+      "loss": 3.9765,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.214417212681387e-05,
+      "loss": 3.9632,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.213578617930335e-05,
+      "loss": 3.9715,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.212740023179283e-05,
+      "loss": 3.9781,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2119047041889774e-05,
+      "loss": 3.9706,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2110661094379254e-05,
+      "loss": 3.9749,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2102275146868734e-05,
+      "loss": 3.9425,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2093889199358214e-05,
+      "loss": 3.9672,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.208551963065143e-05,
+      "loss": 3.9782,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20771336831409e-05,
+      "loss": 3.9726,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206874773563038e-05,
+      "loss": 3.969,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206036178811986e-05,
+      "loss": 3.9596,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205197584060934e-05,
+      "loss": 3.9605,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.204360627190255e-05,
+      "loss": 3.9486,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203522032439203e-05,
+      "loss": 3.9757,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202683437688151e-05,
+      "loss": 3.9491,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201844842937099e-05,
+      "loss": 3.9732,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201006248186048e-05,
+      "loss": 3.9686,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.200167653434996e-05,
+      "loss": 3.9546,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.199329058683944e-05,
+      "loss": 3.9698,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.198490463932892e-05,
+      "loss": 3.9539,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19765186918184e-05,
+      "loss": 3.9569,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196814912311161e-05,
+      "loss": 3.9587,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.195976317560109e-05,
+      "loss": 3.9696,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.195137722809057e-05,
+      "loss": 3.9548,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.194299128058005e-05,
+      "loss": 3.9631,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.193462171187326e-05,
+      "loss": 3.9501,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.192623576436274e-05,
+      "loss": 3.9584,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191784981685222e-05,
+      "loss": 3.9737,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1909463869341697e-05,
+      "loss": 3.9692,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1901077921831177e-05,
+      "loss": 3.96,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1892691974320657e-05,
+      "loss": 3.9674,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188430602681014e-05,
+      "loss": 3.9793,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187593645810335e-05,
+      "loss": 3.9552,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186755051059283e-05,
+      "loss": 3.9692,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185916456308231e-05,
+      "loss": 3.9587,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185077861557179e-05,
+      "loss": 3.9623,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1842409046865e-05,
+      "loss": 3.9601,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.183402309935448e-05,
+      "loss": 3.9627,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182563715184396e-05,
+      "loss": 3.9768,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.181725120433344e-05,
+      "loss": 3.9734,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1808865256822914e-05,
+      "loss": 3.9607,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1800479309312394e-05,
+      "loss": 3.9426,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.179209336180188e-05,
+      "loss": 3.9551,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.178370741429136e-05,
+      "loss": 3.9674,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.177533784558457e-05,
+      "loss": 3.9604,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176695189807405e-05,
+      "loss": 3.9576,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.175856595056353e-05,
+      "loss": 3.9658,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.175019638185674e-05,
+      "loss": 3.9625,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174181043434622e-05,
+      "loss": 3.9643,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.17334244868357e-05,
+      "loss": 3.955,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172503853932518e-05,
+      "loss": 3.9669,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.171665259181466e-05,
+      "loss": 3.9571,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170826664430414e-05,
+      "loss": 3.974,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169989707559735e-05,
+      "loss": 3.9521,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1691511128086835e-05,
+      "loss": 3.9543,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1683125180576315e-05,
+      "loss": 3.9672,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1674739233065795e-05,
+      "loss": 3.961,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1666353285555275e-05,
+      "loss": 3.9699,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1657967338044755e-05,
+      "loss": 3.9598,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1649581390534234e-05,
+      "loss": 3.96,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1641195443023714e-05,
+      "loss": 3.9556,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1632825874316924e-05,
+      "loss": 3.9588,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1624439926806403e-05,
+      "loss": 3.9521,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161607035809961e-05,
+      "loss": 3.9583,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160768441058909e-05,
+      "loss": 3.9603,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159929846307857e-05,
+      "loss": 3.9559,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159091251556805e-05,
+      "loss": 3.9553,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.158252656805753e-05,
+      "loss": 3.9617,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157414062054702e-05,
+      "loss": 3.9447,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15657546730365e-05,
+      "loss": 3.9673,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155736872552598e-05,
+      "loss": 3.9605,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154898277801545e-05,
+      "loss": 3.9652,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154061320930867e-05,
+      "loss": 3.9678,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153222726179815e-05,
+      "loss": 3.9703,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152385769309136e-05,
+      "loss": 3.9601,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.151547174558084e-05,
+      "loss": 3.9554,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150708579807032e-05,
+      "loss": 3.9632,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14986998505598e-05,
+      "loss": 3.9618,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149031390304928e-05,
+      "loss": 3.9516,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.148192795553876e-05,
+      "loss": 3.9715,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.147354200802824e-05,
+      "loss": 3.9553,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.146515606051772e-05,
+      "loss": 3.9645,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1456786491810926e-05,
+      "loss": 3.9528,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1448400544300406e-05,
+      "loss": 3.9511,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1440014596789886e-05,
+      "loss": 3.967,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14316450280831e-05,
+      "loss": 3.9562,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1423259080572575e-05,
+      "loss": 3.9594,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1414873133062055e-05,
+      "loss": 3.9723,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1406487185551535e-05,
+      "loss": 3.9579,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1398101238041015e-05,
+      "loss": 3.9528,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1389715290530495e-05,
+      "loss": 3.9649,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1381329343019975e-05,
+      "loss": 3.9651,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1372943395509455e-05,
+      "loss": 3.9648,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.136457382680267e-05,
+      "loss": 3.9542,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135618787929215e-05,
+      "loss": 3.9579,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134780193178163e-05,
+      "loss": 3.9728,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133941598427111e-05,
+      "loss": 3.9736,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133103003676059e-05,
+      "loss": 3.9575,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13226604680538e-05,
+      "loss": 3.9624,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.131429089934701e-05,
+      "loss": 3.9691,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.130590495183649e-05,
+      "loss": 3.9611,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.12975353831297e-05,
+      "loss": 3.9617,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128914943561918e-05,
+      "loss": 3.9589,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128076348810866e-05,
+      "loss": 3.9562,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1272377540598144e-05,
+      "loss": 3.9686,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1263991593087624e-05,
+      "loss": 3.9518,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.026181697845459,
+      "eval_runtime": 292.8384,
+      "eval_samples_per_second": 1303.077,
+      "eval_steps_per_second": 40.722,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1255605645577104e-05,
+      "loss": 3.9386,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1247219698066584e-05,
+      "loss": 3.9524,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1238833750556064e-05,
+      "loss": 3.9531,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1230447803045544e-05,
+      "loss": 3.9739,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1222061855535024e-05,
+      "loss": 3.9704,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1213675908024504e-05,
+      "loss": 3.9558,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1205289960513984e-05,
+      "loss": 3.9637,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1196904013003464e-05,
+      "loss": 3.9579,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118851806549294e-05,
+      "loss": 3.9559,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118013211798242e-05,
+      "loss": 3.9665,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11717461704719e-05,
+      "loss": 3.9603,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116336022296138e-05,
+      "loss": 3.965,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.115499065425459e-05,
+      "loss": 3.9722,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114660470674407e-05,
+      "loss": 3.9536,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113823513803729e-05,
+      "loss": 3.9504,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112984919052676e-05,
+      "loss": 3.9429,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112146324301624e-05,
+      "loss": 3.96,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111307729550572e-05,
+      "loss": 3.9593,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11046913479952e-05,
+      "loss": 3.9543,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109630540048468e-05,
+      "loss": 3.9582,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108791945297416e-05,
+      "loss": 3.9738,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107953350546364e-05,
+      "loss": 3.9628,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107114755795312e-05,
+      "loss": 3.9653,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.10627616104426e-05,
+      "loss": 3.9594,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.105439204173581e-05,
+      "loss": 3.9688,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.10460060942253e-05,
+      "loss": 3.9552,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103762014671478e-05,
+      "loss": 3.9579,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102923419920426e-05,
+      "loss": 3.9528,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102084825169374e-05,
+      "loss": 3.9517,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101246230418322e-05,
+      "loss": 3.9551,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.10040763566727e-05,
+      "loss": 3.9542,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099569040916218e-05,
+      "loss": 3.9582,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0987320840455386e-05,
+      "loss": 3.9631,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0978934892944866e-05,
+      "loss": 3.9605,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0970548945434346e-05,
+      "loss": 3.963,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0962179376727555e-05,
+      "loss": 3.9609,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0953793429217035e-05,
+      "loss": 3.9642,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0945407481706515e-05,
+      "loss": 3.94,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0937021534195995e-05,
+      "loss": 3.9654,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0928635586685475e-05,
+      "loss": 3.9389,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0920249639174955e-05,
+      "loss": 3.9554,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091188007046817e-05,
+      "loss": 3.9555,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.090351050176138e-05,
+      "loss": 3.9646,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089512455425086e-05,
+      "loss": 3.9483,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088673860674034e-05,
+      "loss": 3.9628,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087835265922982e-05,
+      "loss": 3.964,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08699667117193e-05,
+      "loss": 3.961,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086158076420877e-05,
+      "loss": 3.9593,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085319481669825e-05,
+      "loss": 3.9358,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084480886918773e-05,
+      "loss": 3.9534,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.083643930048095e-05,
+      "loss": 3.9673,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082805335297043e-05,
+      "loss": 3.9594,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081966740545991e-05,
+      "loss": 3.9555,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081128145794939e-05,
+      "loss": 3.9472,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.080289551043887e-05,
+      "loss": 3.9469,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.079452594173208e-05,
+      "loss": 3.9373,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.078613999422156e-05,
+      "loss": 3.9652,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077775404671104e-05,
+      "loss": 3.9353,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076936809920052e-05,
+      "loss": 3.9585,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076098215169e-05,
+      "loss": 3.9587,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075259620417948e-05,
+      "loss": 3.9447,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074421025666896e-05,
+      "loss": 3.9531,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0735840687962166e-05,
+      "loss": 3.9421,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072745474045165e-05,
+      "loss": 3.9461,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071908517174486e-05,
+      "loss": 3.9409,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071069922423434e-05,
+      "loss": 3.9615,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070231327672382e-05,
+      "loss": 3.9466,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06939273292133e-05,
+      "loss": 3.9445,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.068554138170278e-05,
+      "loss": 3.9416,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067715543419226e-05,
+      "loss": 3.948,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066876948668174e-05,
+      "loss": 3.9621,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066038353917122e-05,
+      "loss": 3.9542,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06519975916607e-05,
+      "loss": 3.9453,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.064361164415018e-05,
+      "loss": 3.9562,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.063522569663966e-05,
+      "loss": 3.9672,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0626839749129135e-05,
+      "loss": 3.943,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061845380161862e-05,
+      "loss": 3.9604,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061008423291184e-05,
+      "loss": 3.9473,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060169828540131e-05,
+      "loss": 3.9496,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059331233789079e-05,
+      "loss": 3.9462,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058492639038027e-05,
+      "loss": 3.9521,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057654044286975e-05,
+      "loss": 3.9635,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056817087416296e-05,
+      "loss": 3.9611,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0559801305456176e-05,
+      "loss": 3.9495,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0551415357945655e-05,
+      "loss": 3.929,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054302941043513e-05,
+      "loss": 3.9483,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053464346292461e-05,
+      "loss": 3.9509,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052625751541409e-05,
+      "loss": 3.9527,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0517871567903575e-05,
+      "loss": 3.9413,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509485620393052e-05,
+      "loss": 3.9551,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0501099672882532e-05,
+      "loss": 3.9521,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0492713725372012e-05,
+      "loss": 3.9534,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0484327777861492e-05,
+      "loss": 3.9402,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0475958209154704e-05,
+      "loss": 3.9543,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0467572261644184e-05,
+      "loss": 3.9454,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0459186314133664e-05,
+      "loss": 3.9636,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0450800366623144e-05,
+      "loss": 3.9464,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442414419112624e-05,
+      "loss": 3.9377,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0434028471602104e-05,
+      "loss": 3.9512,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0425642524091584e-05,
+      "loss": 3.951,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0417256576581067e-05,
+      "loss": 3.9573,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0408870629070547e-05,
+      "loss": 3.9479,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0400501060363756e-05,
+      "loss": 3.9484,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0392115112853236e-05,
+      "loss": 3.9421,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0383745544146445e-05,
+      "loss": 3.9495,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.037535959663593e-05,
+      "loss": 3.9376,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036697364912541e-05,
+      "loss": 3.9478,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035858770161489e-05,
+      "loss": 3.9496,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035020175410437e-05,
+      "loss": 3.9478,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.034181580659385e-05,
+      "loss": 3.9382,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0333446237887058e-05,
+      "loss": 3.9501,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032507666918027e-05,
+      "loss": 3.9381,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.031669072166975e-05,
+      "loss": 3.9519,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.030830477415923e-05,
+      "loss": 3.9487,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.029991882664871e-05,
+      "loss": 3.9485,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.029153287913819e-05,
+      "loss": 3.9574,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.028314693162767e-05,
+      "loss": 3.9553,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0274760984117147e-05,
+      "loss": 3.9517,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0266375036606626e-05,
+      "loss": 3.9403,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0257989089096106e-05,
+      "loss": 3.9515,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0249603141585586e-05,
+      "loss": 3.951,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0241233572878795e-05,
+      "loss": 3.9414,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0232847625368275e-05,
+      "loss": 3.9547,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.022446167785776e-05,
+      "loss": 3.9454,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.021607573034724e-05,
+      "loss": 3.9541,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.020768978283672e-05,
+      "loss": 3.9391,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.01993038353262e-05,
+      "loss": 3.9426,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.019091788781568e-05,
+      "loss": 3.9486,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.018253194030516e-05,
+      "loss": 3.9454,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174162371598368e-05,
+      "loss": 3.942,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.016577642408785e-05,
+      "loss": 3.9633,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.015739047657733e-05,
+      "loss": 3.949,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014902090787054e-05,
+      "loss": 3.9431,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014063496036002e-05,
+      "loss": 3.9512,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.01322490128495e-05,
+      "loss": 3.9525,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.012386306533898e-05,
+      "loss": 3.9526,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.011547711782846e-05,
+      "loss": 3.9478,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010709117031794e-05,
+      "loss": 3.9432,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0098705222807423e-05,
+      "loss": 3.9602,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0090319275296903e-05,
+      "loss": 3.9608,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0081949706590112e-05,
+      "loss": 3.9461,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0073563759079592e-05,
+      "loss": 3.9487,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0065177811569072e-05,
+      "loss": 3.9595,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0056791864058552e-05,
+      "loss": 3.9459,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0048405916548032e-05,
+      "loss": 3.9511,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004001996903751e-05,
+      "loss": 3.948,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0031650400330724e-05,
+      "loss": 3.9457,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023264452820204e-05,
+      "loss": 3.9582,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0014878505309678e-05,
+      "loss": 3.94,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.000649255779916e-05,
+      "loss": 3.9451,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.019927024841309,
+      "eval_runtime": 283.2659,
+      "eval_samples_per_second": 1347.112,
+      "eval_steps_per_second": 42.098,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.999810661028864e-05,
+      "loss": 3.9399,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.998972066277812e-05,
+      "loss": 3.9398,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.99813347152676e-05,
+      "loss": 3.962,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.997294876775708e-05,
+      "loss": 3.9593,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.996457919905029e-05,
+      "loss": 3.947,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.995619325153977e-05,
+      "loss": 3.9541,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9947823682832982e-05,
+      "loss": 3.9467,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9939437735322462e-05,
+      "loss": 3.938,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9931051787811942e-05,
+      "loss": 3.958,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9922665840301422e-05,
+      "loss": 3.9476,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9914279892790902e-05,
+      "loss": 3.9569,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9905893945280382e-05,
+      "loss": 3.9537,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9897524376573595e-05,
+      "loss": 3.9476,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9889138429063075e-05,
+      "loss": 3.9375,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9880752481552554e-05,
+      "loss": 3.9303,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9872366534042034e-05,
+      "loss": 3.9444,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9863980586531514e-05,
+      "loss": 3.9471,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9855594639020994e-05,
+      "loss": 3.945,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9847208691510474e-05,
+      "loss": 3.9441,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9838839122803687e-05,
+      "loss": 3.967,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9830469554096896e-05,
+      "loss": 3.9487,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9822083606586376e-05,
+      "loss": 3.9525,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9813697659075856e-05,
+      "loss": 3.9473,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9805311711565336e-05,
+      "loss": 3.9624,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9796925764054816e-05,
+      "loss": 3.9374,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.97885398165443e-05,
+      "loss": 3.9475,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.978015386903378e-05,
+      "loss": 3.9417,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9771784300326988e-05,
+      "loss": 3.9447,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9763398352816468e-05,
+      "loss": 3.9439,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9755012405305948e-05,
+      "loss": 3.942,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9746626457795428e-05,
+      "loss": 3.9464,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9738240510284908e-05,
+      "loss": 3.9532,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.972985456277439e-05,
+      "loss": 3.9469,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.972146861526387e-05,
+      "loss": 3.9536,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.971309904655708e-05,
+      "loss": 3.948,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.970471309904656e-05,
+      "loss": 3.9549,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.969632715153604e-05,
+      "loss": 3.9311,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9687941204025517e-05,
+      "loss": 3.953,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9679555256514997e-05,
+      "loss": 3.9294,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9671169309004477e-05,
+      "loss": 3.9464,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9662799740297693e-05,
+      "loss": 3.943,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9654413792787166e-05,
+      "loss": 3.9521,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9646027845276646e-05,
+      "loss": 3.9359,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.963764189776613e-05,
+      "loss": 3.9541,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962925595025561e-05,
+      "loss": 3.9506,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962087000274509e-05,
+      "loss": 3.95,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.961248405523457e-05,
+      "loss": 3.9506,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9604114486527778e-05,
+      "loss": 3.9254,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9595728539017258e-05,
+      "loss": 3.9443,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.958735897031047e-05,
+      "loss": 3.9547,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957897302279995e-05,
+      "loss": 3.9463,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957058707528943e-05,
+      "loss": 3.947,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.956220112777891e-05,
+      "loss": 3.9351,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.955381518026839e-05,
+      "loss": 3.9404,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.954542923275787e-05,
+      "loss": 3.9236,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.953704328524735e-05,
+      "loss": 3.9542,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.952865733773683e-05,
+      "loss": 3.9243,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9520287769030043e-05,
+      "loss": 3.9519,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9511901821519523e-05,
+      "loss": 3.9462,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9503515874009003e-05,
+      "loss": 3.9337,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9495129926498483e-05,
+      "loss": 3.9393,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9486743978987962e-05,
+      "loss": 3.9378,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9478358031477442e-05,
+      "loss": 3.9343,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9469988462770655e-05,
+      "loss": 3.9304,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9461602515260135e-05,
+      "loss": 3.9483,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9453216567749615e-05,
+      "loss": 3.9375,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9444830620239095e-05,
+      "loss": 3.932,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9436444672728575e-05,
+      "loss": 3.9311,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9428058725218055e-05,
+      "loss": 3.9348,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9419689156511264e-05,
+      "loss": 3.9495,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9411303209000747e-05,
+      "loss": 3.947,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9402917261490227e-05,
+      "loss": 3.9346,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.93945313139797e-05,
+      "loss": 3.9461,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.938614536646918e-05,
+      "loss": 3.9539,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9377775797762396e-05,
+      "loss": 3.932,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9369389850251876e-05,
+      "loss": 3.945,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9361003902741353e-05,
+      "loss": 3.941,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9352617955230833e-05,
+      "loss": 3.9405,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.934424838652405e-05,
+      "loss": 3.9359,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.933586243901353e-05,
+      "loss": 3.9408,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9327476491503e-05,
+      "loss": 3.9508,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9319090543992485e-05,
+      "loss": 3.9573,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9310704596481965e-05,
+      "loss": 3.934,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9302318648971445e-05,
+      "loss": 3.9168,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9293932701460925e-05,
+      "loss": 3.9378,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9285546753950405e-05,
+      "loss": 3.9407,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9277177185243614e-05,
+      "loss": 3.9432,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9268791237733094e-05,
+      "loss": 3.936,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9260405290222577e-05,
+      "loss": 3.9429,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9252035721515786e-05,
+      "loss": 3.9411,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9243649774005266e-05,
+      "loss": 3.9397,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9235263826494746e-05,
+      "loss": 3.9307,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9226877878984226e-05,
+      "loss": 3.9438,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9218491931473706e-05,
+      "loss": 3.9354,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.921012236276692e-05,
+      "loss": 3.9505,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.92017364152564e-05,
+      "loss": 3.9404,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.919335046774588e-05,
+      "loss": 3.9258,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.918496452023536e-05,
+      "loss": 3.9398,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.917657857272484e-05,
+      "loss": 3.9403,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.916819262521432e-05,
+      "loss": 3.95,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9159806677703798e-05,
+      "loss": 3.9362,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9151420730193278e-05,
+      "loss": 3.9397,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.914305116148649e-05,
+      "loss": 3.9288,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.913466521397597e-05,
+      "loss": 3.9422,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.912629564526918e-05,
+      "loss": 3.9252,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.911790969775866e-05,
+      "loss": 3.9364,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.910952375024814e-05,
+      "loss": 3.9391,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9101137802737623e-05,
+      "loss": 3.9392,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9092751855227103e-05,
+      "loss": 3.9273,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9084365907716583e-05,
+      "loss": 3.9391,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9075979960206063e-05,
+      "loss": 3.9323,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9067610391499272e-05,
+      "loss": 3.9427,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9059224443988752e-05,
+      "loss": 3.9356,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9050838496478232e-05,
+      "loss": 3.9352,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9042452548967712e-05,
+      "loss": 3.9497,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.903406660145719e-05,
+      "loss": 3.9456,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.902568065394667e-05,
+      "loss": 3.9407,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.901729470643615e-05,
+      "loss": 3.9308,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9008908758925628e-05,
+      "loss": 3.9385,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.900053919021884e-05,
+      "loss": 3.9365,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.899215324270832e-05,
+      "loss": 3.9347,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.89837672951978e-05,
+      "loss": 3.9486,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.897538134768728e-05,
+      "loss": 3.9324,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.896701177898049e-05,
+      "loss": 3.9445,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.895862583146997e-05,
+      "loss": 3.9281,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8950256262763186e-05,
+      "loss": 3.9343,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8941870315252662e-05,
+      "loss": 3.9391,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8933484367742142e-05,
+      "loss": 3.9352,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8925098420231622e-05,
+      "loss": 3.9267,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8916712472721102e-05,
+      "loss": 3.9549,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8908326525210582e-05,
+      "loss": 3.94,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8899956956503794e-05,
+      "loss": 3.9359,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8891571008993274e-05,
+      "loss": 3.9366,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8883185061482754e-05,
+      "loss": 3.9379,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8874799113972234e-05,
+      "loss": 3.9437,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8866413166461714e-05,
+      "loss": 3.9405,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8858027218951194e-05,
+      "loss": 3.9339,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8849657650244403e-05,
+      "loss": 3.9503,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8841271702733887e-05,
+      "loss": 3.944,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8832885755223367e-05,
+      "loss": 3.9401,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8824499807712847e-05,
+      "loss": 3.9401,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8816113860202327e-05,
+      "loss": 3.9439,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8807744291495536e-05,
+      "loss": 3.939,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8799358343985016e-05,
+      "loss": 3.9412,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8790972396474496e-05,
+      "loss": 3.9397,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.878258644896398e-05,
+      "loss": 3.9382,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.877420050145346e-05,
+      "loss": 3.945,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8765830932746668e-05,
+      "loss": 3.931,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8757444985236148e-05,
+      "loss": 3.9351,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.015511512756348,
+      "eval_runtime": 287.5044,
+      "eval_samples_per_second": 1327.253,
+      "eval_steps_per_second": 41.478,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8749059037725628e-05,
+      "loss": 3.9277,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8740673090215108e-05,
+      "loss": 3.9307,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8732287142704588e-05,
+      "loss": 3.9534,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.872390119519407e-05,
+      "loss": 3.9443,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.871551524768355e-05,
+      "loss": 3.9389,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8707129300173024e-05,
+      "loss": 3.9397,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8698743352662504e-05,
+      "loss": 3.939,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8690357405151984e-05,
+      "loss": 3.9294,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8681971457641464e-05,
+      "loss": 3.9451,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8673585510130947e-05,
+      "loss": 3.942,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8665199562620427e-05,
+      "loss": 3.9433,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8656813615109907e-05,
+      "loss": 3.9452,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8648444046403116e-05,
+      "loss": 3.9355,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8640058098892596e-05,
+      "loss": 3.93,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8631672151382076e-05,
+      "loss": 3.9269,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8623286203871556e-05,
+      "loss": 3.9272,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.861490025636104e-05,
+      "loss": 3.9386,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.860653068765425e-05,
+      "loss": 3.9323,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.859814474014373e-05,
+      "loss": 3.9378,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858975879263321e-05,
+      "loss": 3.9588,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858137284512269e-05,
+      "loss": 3.9403,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.857298689761217e-05,
+      "loss": 3.945,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.856460095010165e-05,
+      "loss": 3.9367,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.855621500259113e-05,
+      "loss": 3.9489,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8547829055080612e-05,
+      "loss": 3.9315,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8539443107570085e-05,
+      "loss": 3.9366,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.85310735388633e-05,
+      "loss": 3.9312,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.852268759135278e-05,
+      "loss": 3.9351,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.851430164384226e-05,
+      "loss": 3.9327,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8505915696331737e-05,
+      "loss": 3.9311,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8497529748821217e-05,
+      "loss": 3.9353,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8489160180114433e-05,
+      "loss": 3.9466,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8480774232603913e-05,
+      "loss": 3.9367,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8472388285093386e-05,
+      "loss": 3.9407,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8464018716386602e-05,
+      "loss": 3.9344,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8455632768876082e-05,
+      "loss": 3.9444,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.844724682136556e-05,
+      "loss": 3.9241,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843886087385504e-05,
+      "loss": 3.9403,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843047492634452e-05,
+      "loss": 3.9188,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8422088978834e-05,
+      "loss": 3.9407,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.841370303132348e-05,
+      "loss": 3.9299,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.840531708381296e-05,
+      "loss": 3.9447,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.839694751510617e-05,
+      "loss": 3.9253,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.838856156759565e-05,
+      "loss": 3.9424,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.838017562008513e-05,
+      "loss": 3.941,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.837178967257461e-05,
+      "loss": 3.9394,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.836340372506409e-05,
+      "loss": 3.9433,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.835501777755357e-05,
+      "loss": 3.9159,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.834663183004305e-05,
+      "loss": 3.9273,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8338262261336263e-05,
+      "loss": 3.944,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8329876313825743e-05,
+      "loss": 3.9383,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8321490366315223e-05,
+      "loss": 3.9338,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8313104418804703e-05,
+      "loss": 3.9301,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8304734850097912e-05,
+      "loss": 3.9291,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8296348902587396e-05,
+      "loss": 3.9207,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8287962955076875e-05,
+      "loss": 3.9367,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8279577007566355e-05,
+      "loss": 3.9205,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8271191060055835e-05,
+      "loss": 3.9341,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8262805112545315e-05,
+      "loss": 3.9415,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8254419165034795e-05,
+      "loss": 3.9256,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8246033217524272e-05,
+      "loss": 3.9263,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8237663648817488e-05,
+      "loss": 3.9293,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8229277701306968e-05,
+      "loss": 3.9203,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8220908132600177e-05,
+      "loss": 3.9223,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8212522185089657e-05,
+      "loss": 3.9399,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8204152616382866e-05,
+      "loss": 3.9295,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.819576666887235e-05,
+      "loss": 3.9232,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.818738072136183e-05,
+      "loss": 3.9225,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.817899477385131e-05,
+      "loss": 3.9233,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.817060882634079e-05,
+      "loss": 3.9382,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.816222287883027e-05,
+      "loss": 3.9356,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8153853310123478e-05,
+      "loss": 3.9283,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8145467362612958e-05,
+      "loss": 3.9371,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.813708141510244e-05,
+      "loss": 3.9407,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812869546759192e-05,
+      "loss": 3.9273,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8120309520081395e-05,
+      "loss": 3.9333,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8111923572570874e-05,
+      "loss": 3.9304,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8103537625060354e-05,
+      "loss": 3.9309,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8095151677549834e-05,
+      "loss": 3.9276,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8086765730039318e-05,
+      "loss": 3.9293,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8078379782528798e-05,
+      "loss": 3.9418,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8069993835018278e-05,
+      "loss": 3.9394,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8061624266311487e-05,
+      "loss": 3.9314,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8053238318800967e-05,
+      "loss": 3.9066,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.804486875009418e-05,
+      "loss": 3.925,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.803648280258366e-05,
+      "loss": 3.9321,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.802809685507314e-05,
+      "loss": 3.9316,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.801971090756262e-05,
+      "loss": 3.9302,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.80113249600521e-05,
+      "loss": 3.9281,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8002955391345308e-05,
+      "loss": 3.9327,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7994569443834788e-05,
+      "loss": 3.9268,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.798618349632427e-05,
+      "loss": 3.9204,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.797781392761748e-05,
+      "loss": 3.9367,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.796942798010696e-05,
+      "loss": 3.9215,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.796104203259644e-05,
+      "loss": 3.9408,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.795265608508592e-05,
+      "loss": 3.9349,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.79442701375754e-05,
+      "loss": 3.9143,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.793588419006488e-05,
+      "loss": 3.931,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.792749824255436e-05,
+      "loss": 3.9291,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7919112295043844e-05,
+      "loss": 3.941,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7910726347533324e-05,
+      "loss": 3.9292,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7902340400022803e-05,
+      "loss": 3.9302,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7893954452512283e-05,
+      "loss": 3.9205,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7885584883805493e-05,
+      "loss": 3.9337,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7877198936294972e-05,
+      "loss": 3.9175,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7868812988784452e-05,
+      "loss": 3.9229,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7860427041273936e-05,
+      "loss": 3.9315,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7852057472567145e-05,
+      "loss": 3.9276,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7843671525056625e-05,
+      "loss": 3.9162,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7835285577546105e-05,
+      "loss": 3.9307,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.782689963003558e-05,
+      "loss": 3.924,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.781851368252506e-05,
+      "loss": 3.9292,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.781012773501454e-05,
+      "loss": 3.927,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.780174178750402e-05,
+      "loss": 3.9234,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.77933558399935e-05,
+      "loss": 3.9482,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.778496989248298e-05,
+      "loss": 3.9367,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.777658394497246e-05,
+      "loss": 3.931,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.776819799746194e-05,
+      "loss": 3.9209,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.775981204995142e-05,
+      "loss": 3.929,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7751442481244633e-05,
+      "loss": 3.9294,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7743072912537843e-05,
+      "loss": 3.9223,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7734686965027323e-05,
+      "loss": 3.94,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7726301017516803e-05,
+      "loss": 3.9281,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7717915070006282e-05,
+      "loss": 3.9277,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7709529122495766e-05,
+      "loss": 3.9219,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7701143174985246e-05,
+      "loss": 3.9247,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7692757227474726e-05,
+      "loss": 3.9254,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7684371279964206e-05,
+      "loss": 3.928,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7676001711257415e-05,
+      "loss": 3.9149,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7667615763746895e-05,
+      "loss": 3.9462,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7659246195040107e-05,
+      "loss": 3.9365,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7650860247529587e-05,
+      "loss": 3.9264,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7642474300019067e-05,
+      "loss": 3.927,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7634088352508547e-05,
+      "loss": 3.9243,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7625702404998027e-05,
+      "loss": 3.9372,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7617316457487507e-05,
+      "loss": 3.9315,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7608930509976987e-05,
+      "loss": 3.9236,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7600544562466467e-05,
+      "loss": 3.9399,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7592158614955943e-05,
+      "loss": 3.9309,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.758378904624916e-05,
+      "loss": 3.934,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.757540309873864e-05,
+      "loss": 3.9285,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.756701715122812e-05,
+      "loss": 3.9373,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.755864758252133e-05,
+      "loss": 3.9324,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.755026163501081e-05,
+      "loss": 3.928,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.754187568750029e-05,
+      "loss": 3.9302,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7533489739989765e-05,
+      "loss": 3.9322,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.752512017128298e-05,
+      "loss": 3.9339,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.751673422377246e-05,
+      "loss": 3.9223,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.750834827626194e-05,
+      "loss": 3.927,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.011007308959961,
+      "eval_runtime": 293.5076,
+      "eval_samples_per_second": 1300.106,
+      "eval_steps_per_second": 40.629,
+      "step": 1373760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7499962328751417e-05,
+      "loss": 3.9202,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7491576381240897e-05,
+      "loss": 3.9221,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7483190433730377e-05,
+      "loss": 3.9432,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7474804486219857e-05,
+      "loss": 3.9389,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7466418538709337e-05,
+      "loss": 3.9286,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7458032591198817e-05,
+      "loss": 3.9296,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7449646643688297e-05,
+      "loss": 3.9304,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7441260696177777e-05,
+      "loss": 3.9173,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.743289112747099e-05,
+      "loss": 3.9383,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.742450517996047e-05,
+      "loss": 3.9324,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.741611923244995e-05,
+      "loss": 3.9321,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.740773328493943e-05,
+      "loss": 3.9352,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739936371623264e-05,
+      "loss": 3.9236,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739097776872212e-05,
+      "loss": 3.9244,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.73825918212116e-05,
+      "loss": 3.9206,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.737420587370108e-05,
+      "loss": 3.9219,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.736581992619056e-05,
+      "loss": 3.9259,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.735745035748377e-05,
+      "loss": 3.9236,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.734906440997325e-05,
+      "loss": 3.9296,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.734067846246273e-05,
+      "loss": 3.954,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7332292514952214e-05,
+      "loss": 3.9291,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7323922946245423e-05,
+      "loss": 3.9325,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7315536998734903e-05,
+      "loss": 3.931,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7307151051224383e-05,
+      "loss": 3.9379,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7298765103713863e-05,
+      "loss": 3.9201,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7290379156203343e-05,
+      "loss": 3.9309,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7281993208692823e-05,
+      "loss": 3.9206,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7273623639986035e-05,
+      "loss": 3.9235,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7265237692475515e-05,
+      "loss": 3.9225,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7256851744964995e-05,
+      "loss": 3.9216,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7248465797454475e-05,
+      "loss": 3.9269,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7240079849943955e-05,
+      "loss": 3.9376,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7231710281237168e-05,
+      "loss": 3.9272,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7223324333726647e-05,
+      "loss": 3.9335,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7214954765019857e-05,
+      "loss": 3.9194,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7206568817509337e-05,
+      "loss": 3.9401,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7198182869998817e-05,
+      "loss": 3.9122,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7189796922488296e-05,
+      "loss": 3.932,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7181410974977776e-05,
+      "loss": 3.9098,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7173025027467253e-05,
+      "loss": 3.9292,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7164639079956733e-05,
+      "loss": 3.9202,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7156253132446213e-05,
+      "loss": 3.9362,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7147867184935693e-05,
+      "loss": 3.9188,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7139481237425173e-05,
+      "loss": 3.9289,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7131095289914653e-05,
+      "loss": 3.9333,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7122725721207865e-05,
+      "loss": 3.933,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7114339773697345e-05,
+      "loss": 3.9307,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7105953826186825e-05,
+      "loss": 3.9122,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7097567878676305e-05,
+      "loss": 3.9151,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7089198309969514e-05,
+      "loss": 3.9376,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7080812362458998e-05,
+      "loss": 3.929,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7072442793752207e-05,
+      "loss": 3.9244,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7064056846241687e-05,
+      "loss": 3.9211,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7055670898731167e-05,
+      "loss": 3.9225,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7047284951220647e-05,
+      "loss": 3.9093,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.703891538251386e-05,
+      "loss": 3.9251,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.703052943500334e-05,
+      "loss": 3.9117,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.702214348749282e-05,
+      "loss": 3.924,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.70137575399823e-05,
+      "loss": 3.9321,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.700537159247178e-05,
+      "loss": 3.9147,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.699698564496126e-05,
+      "loss": 3.9153,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698859969745074e-05,
+      "loss": 3.924,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698021374994022e-05,
+      "loss": 3.909,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.69718278024297e-05,
+      "loss": 3.911,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.696344185491918e-05,
+      "loss": 3.9303,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.695507228621239e-05,
+      "loss": 3.922,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.694668633870187e-05,
+      "loss": 3.9126,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.693830039119135e-05,
+      "loss": 3.9129,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.692991444368083e-05,
+      "loss": 3.9173,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.692152849617031e-05,
+      "loss": 3.9297,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6913142548659787e-05,
+      "loss": 3.9266,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6904756601149267e-05,
+      "loss": 3.9184,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6896387032442483e-05,
+      "loss": 3.9299,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6888001084931963e-05,
+      "loss": 3.9275,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6879615137421436e-05,
+      "loss": 3.919,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.687122918991092e-05,
+      "loss": 3.9223,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.68628432424004e-05,
+      "loss": 3.9247,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.685445729488988e-05,
+      "loss": 3.9221,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.684607134737936e-05,
+      "loss": 3.9205,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.683768539986884e-05,
+      "loss": 3.918,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.682931583116205e-05,
+      "loss": 3.93,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.682094626245526e-05,
+      "loss": 3.9337,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.681256031494474e-05,
+      "loss": 3.9183,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.680417436743422e-05,
+      "loss": 3.9039,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.67957884199237e-05,
+      "loss": 3.9132,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.678740247241318e-05,
+      "loss": 3.9214,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.677901652490266e-05,
+      "loss": 3.9271,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.677064695619587e-05,
+      "loss": 3.9202,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6762261008685353e-05,
+      "loss": 3.917,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6753875061174833e-05,
+      "loss": 3.9291,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6745489113664313e-05,
+      "loss": 3.9135,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6737103166153793e-05,
+      "loss": 3.9143,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6728733597447002e-05,
+      "loss": 3.926,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6720347649936482e-05,
+      "loss": 3.9167,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6711961702425962e-05,
+      "loss": 3.9334,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6703575754915446e-05,
+      "loss": 3.9228,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6695189807404926e-05,
+      "loss": 3.9052,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6686803859894406e-05,
+      "loss": 3.9223,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6678417912383885e-05,
+      "loss": 3.9164,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6670031964873365e-05,
+      "loss": 3.9337,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6661646017362845e-05,
+      "loss": 3.9246,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6653276448656054e-05,
+      "loss": 3.9172,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6644890501145538e-05,
+      "loss": 3.9081,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6636504553635018e-05,
+      "loss": 3.9237,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6628118606124498e-05,
+      "loss": 3.9071,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6619732658613978e-05,
+      "loss": 3.9164,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6611363089907187e-05,
+      "loss": 3.9218,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6602977142396667e-05,
+      "loss": 3.9179,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6594591194886147e-05,
+      "loss": 3.9065,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6586205247375623e-05,
+      "loss": 3.9281,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.657783567866884e-05,
+      "loss": 3.913,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.656944973115832e-05,
+      "loss": 3.9167,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.65610637836478e-05,
+      "loss": 3.9219,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6552677836137276e-05,
+      "loss": 3.9126,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.654430826743049e-05,
+      "loss": 3.935,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.653592231991997e-05,
+      "loss": 3.927,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.652756913001691e-05,
+      "loss": 3.9251,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651918318250639e-05,
+      "loss": 3.9086,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651079723499587e-05,
+      "loss": 3.9178,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6502411287485353e-05,
+      "loss": 3.9226,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6494025339974833e-05,
+      "loss": 3.9089,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6485655771268042e-05,
+      "loss": 3.9272,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6477269823757522e-05,
+      "loss": 3.9188,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6468883876247002e-05,
+      "loss": 3.9187,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6460497928736482e-05,
+      "loss": 3.9183,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6452111981225962e-05,
+      "loss": 3.9186,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6443726033715445e-05,
+      "loss": 3.9089,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6435356465008654e-05,
+      "loss": 3.9237,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6426970517498134e-05,
+      "loss": 3.9087,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6418584569987614e-05,
+      "loss": 3.9331,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6410198622477094e-05,
+      "loss": 3.9311,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.640181267496657e-05,
+      "loss": 3.916,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.639342672745605e-05,
+      "loss": 3.9165,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.638504077994553e-05,
+      "loss": 3.917,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.637665483243501e-05,
+      "loss": 3.924,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636828526372822e-05,
+      "loss": 3.924,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.63598993162177e-05,
+      "loss": 3.9167,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6351529747510915e-05,
+      "loss": 3.9288,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6343143800000392e-05,
+      "loss": 3.9243,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6334757852489872e-05,
+      "loss": 3.9252,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6326371904979352e-05,
+      "loss": 3.9157,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6317985957468832e-05,
+      "loss": 3.9312,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6309600009958312e-05,
+      "loss": 3.9243,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6301214062447792e-05,
+      "loss": 3.9165,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6292828114937272e-05,
+      "loss": 3.9248,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6284442167426755e-05,
+      "loss": 3.9234,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6276056219916235e-05,
+      "loss": 3.9207,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6267670272405715e-05,
+      "loss": 3.9118,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6259284324895195e-05,
+      "loss": 3.9139,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.006725788116455,
+      "eval_runtime": 299.0312,
+      "eval_samples_per_second": 1276.091,
+      "eval_steps_per_second": 39.879,
+      "step": 1450080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6250898377384675e-05,
+      "loss": 3.9111,
+      "step": 1450496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6242512429874155e-05,
+      "loss": 3.9117,
+      "step": 1451008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6234126482363635e-05,
+      "loss": 3.9333,
+      "step": 1451520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.622574053485311e-05,
+      "loss": 3.9271,
+      "step": 1452032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.621735458734259e-05,
+      "loss": 3.9279,
+      "step": 1452544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.620896863983207e-05,
+      "loss": 3.9119,
+      "step": 1453056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.620058269232155e-05,
+      "loss": 3.9243,
+      "step": 1453568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.619219674481103e-05,
+      "loss": 3.9066,
+      "step": 1454080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.618381079730051e-05,
+      "loss": 3.931,
+      "step": 1454592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.617542484978999e-05,
+      "loss": 3.925,
+      "step": 1455104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.616703890227947e-05,
+      "loss": 3.9188,
+      "step": 1455616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6158652954768954e-05,
+      "loss": 3.9295,
+      "step": 1456128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6150283386062164e-05,
+      "loss": 3.9147,
+      "step": 1456640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6141897438551644e-05,
+      "loss": 3.9161,
+      "step": 1457152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6133511491041123e-05,
+      "loss": 3.9082,
+      "step": 1457664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6125125543530603e-05,
+      "loss": 3.9169,
+      "step": 1458176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6116739596020083e-05,
+      "loss": 3.9147,
+      "step": 1458688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6108353648509563e-05,
+      "loss": 3.9149,
+      "step": 1459200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6099967700999047e-05,
+      "loss": 3.918,
+      "step": 1459712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6091581753488527e-05,
+      "loss": 3.9424,
+      "step": 1460224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6083212184781736e-05,
+      "loss": 3.9212,
+      "step": 1460736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6074826237271216e-05,
+      "loss": 3.9238,
+      "step": 1461248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6066440289760696e-05,
+      "loss": 3.9214,
+      "step": 1461760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6058054342250172e-05,
+      "loss": 3.9289,
+      "step": 1462272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6049684773543388e-05,
+      "loss": 3.9125,
+      "step": 1462784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6041298826032868e-05,
+      "loss": 3.9217,
+      "step": 1463296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6032912878522348e-05,
+      "loss": 3.9131,
+      "step": 1463808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.602452693101182e-05,
+      "loss": 3.9124,
+      "step": 1464320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.60161409835013e-05,
+      "loss": 3.912,
+      "step": 1464832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6007755035990784e-05,
+      "loss": 3.9171,
+      "step": 1465344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5999369088480264e-05,
+      "loss": 3.9162,
+      "step": 1465856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5990983140969744e-05,
+      "loss": 3.9288,
+      "step": 1466368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5982597193459224e-05,
+      "loss": 3.9192,
+      "step": 1466880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5974211245948704e-05,
+      "loss": 3.9228,
+      "step": 1467392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5965858056045646e-05,
+      "loss": 3.9145,
+      "step": 1467904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5957472108535126e-05,
+      "loss": 3.9268,
+      "step": 1468416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5949086161024606e-05,
+      "loss": 3.9105,
+      "step": 1468928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5940700213514086e-05,
+      "loss": 3.9208,
+      "step": 1469440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5932314266003566e-05,
+      "loss": 3.9014,
+      "step": 1469952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5923928318493046e-05,
+      "loss": 3.9175,
+      "step": 1470464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5915542370982526e-05,
+      "loss": 3.9154,
+      "step": 1470976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5907156423472006e-05,
+      "loss": 3.9296,
+      "step": 1471488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5898770475961486e-05,
+      "loss": 3.911,
+      "step": 1472000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5890384528450966e-05,
+      "loss": 3.9187,
+      "step": 1472512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.588199858094045e-05,
+      "loss": 3.9229,
+      "step": 1473024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.587361263342993e-05,
+      "loss": 3.9254,
+      "step": 1473536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5865243064723138e-05,
+      "loss": 3.9187,
+      "step": 1474048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.585688987482008e-05,
+      "loss": 3.9087,
+      "step": 1474560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.584850392730956e-05,
+      "loss": 3.9015,
+      "step": 1475072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.584011797979904e-05,
+      "loss": 3.9267,
+      "step": 1475584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.583173203228852e-05,
+      "loss": 3.9212,
+      "step": 1476096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5823346084778e-05,
+      "loss": 3.9148,
+      "step": 1476608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.581496013726748e-05,
+      "loss": 3.9105,
+      "step": 1477120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.580657418975696e-05,
+      "loss": 3.9159,
+      "step": 1477632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.579818824224644e-05,
+      "loss": 3.8976,
+      "step": 1478144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.578980229473592e-05,
+      "loss": 3.9194,
+      "step": 1478656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5781416347225403e-05,
+      "loss": 3.9009,
+      "step": 1479168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5773030399714882e-05,
+      "loss": 3.917,
+      "step": 1479680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5764644452204356e-05,
+      "loss": 3.9218,
+      "step": 1480192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.575627488349757e-05,
+      "loss": 3.9119,
+      "step": 1480704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.574790531479078e-05,
+      "loss": 3.9017,
+      "step": 1481216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5739519367280264e-05,
+      "loss": 3.9148,
+      "step": 1481728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5731133419769744e-05,
+      "loss": 3.9016,
+      "step": 1482240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5722747472259224e-05,
+      "loss": 3.9072,
+      "step": 1482752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5714377903552433e-05,
+      "loss": 3.9185,
+      "step": 1483264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5705991956041913e-05,
+      "loss": 3.9134,
+      "step": 1483776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5697606008531393e-05,
+      "loss": 3.905,
+      "step": 1484288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5689220061020873e-05,
+      "loss": 3.9016,
+      "step": 1484800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5680834113510356e-05,
+      "loss": 3.9087,
+      "step": 1485312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5672464544803565e-05,
+      "loss": 3.9149,
+      "step": 1485824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5664078597293045e-05,
+      "loss": 3.9192,
+      "step": 1486336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5655692649782525e-05,
+      "loss": 3.9139,
+      "step": 1486848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5647323081075734e-05,
+      "loss": 3.9196,
+      "step": 1487360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5638937133565218e-05,
+      "loss": 3.9218,
+      "step": 1487872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5630551186054698e-05,
+      "loss": 3.9098,
+      "step": 1488384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5622165238544178e-05,
+      "loss": 3.911,
+      "step": 1488896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5613779291033658e-05,
+      "loss": 3.9164,
+      "step": 1489408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.560539334352313e-05,
+      "loss": 3.9162,
+      "step": 1489920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.559700739601261e-05,
+      "loss": 3.913,
+      "step": 1490432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5588637827305827e-05,
+      "loss": 3.9093,
+      "step": 1490944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5580251879795303e-05,
+      "loss": 3.9192,
+      "step": 1491456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5571865932284783e-05,
+      "loss": 3.9234,
+      "step": 1491968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5563479984774263e-05,
+      "loss": 3.9133,
+      "step": 1492480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5555094037263743e-05,
+      "loss": 3.8986,
+      "step": 1492992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5546708089753223e-05,
+      "loss": 3.9011,
+      "step": 1493504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5538322142242703e-05,
+      "loss": 3.9106,
+      "step": 1494016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5529936194732186e-05,
+      "loss": 3.9181,
+      "step": 1494528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5521550247221666e-05,
+      "loss": 3.9155,
+      "step": 1495040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5513180678514875e-05,
+      "loss": 3.9042,
+      "step": 1495552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5504794731004355e-05,
+      "loss": 3.9186,
+      "step": 1496064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5496408783493835e-05,
+      "loss": 3.9104,
+      "step": 1496576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5488022835983315e-05,
+      "loss": 3.9067,
+      "step": 1497088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5479636888472795e-05,
+      "loss": 3.9104,
+      "step": 1497600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.547125094096228e-05,
+      "loss": 3.9159,
+      "step": 1498112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5462881372255488e-05,
+      "loss": 3.9236,
+      "step": 1498624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5454495424744967e-05,
+      "loss": 3.9112,
+      "step": 1499136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5446109477234447e-05,
+      "loss": 3.8969,
+      "step": 1499648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5437723529723927e-05,
+      "loss": 3.9155,
+      "step": 1500160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5429337582213407e-05,
+      "loss": 3.9081,
+      "step": 1500672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5420951634702887e-05,
+      "loss": 3.9247,
+      "step": 1501184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.54125820659961e-05,
+      "loss": 3.9178,
+      "step": 1501696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.540419611848558e-05,
+      "loss": 3.9086,
+      "step": 1502208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.539581017097506e-05,
+      "loss": 3.9016,
+      "step": 1502720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.538742422346454e-05,
+      "loss": 3.9178,
+      "step": 1503232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.537903827595402e-05,
+      "loss": 3.899,
+      "step": 1503744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5370668707247232e-05,
+      "loss": 3.9097,
+      "step": 1504256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5362282759736712e-05,
+      "loss": 3.9072,
+      "step": 1504768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5353896812226192e-05,
+      "loss": 3.9143,
+      "step": 1505280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5345510864715665e-05,
+      "loss": 3.8978,
+      "step": 1505792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.533714129600888e-05,
+      "loss": 3.9185,
+      "step": 1506304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.532875534849836e-05,
+      "loss": 3.9042,
+      "step": 1506816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.532036940098784e-05,
+      "loss": 3.9065,
+      "step": 1507328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5311983453477318e-05,
+      "loss": 3.9174,
+      "step": 1507840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5303597505966798e-05,
+      "loss": 3.9036,
+      "step": 1508352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5295211558456277e-05,
+      "loss": 3.9289,
+      "step": 1508864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5286825610945757e-05,
+      "loss": 3.9201,
+      "step": 1509376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.527845604223897e-05,
+      "loss": 3.9142,
+      "step": 1509888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.527007009472845e-05,
+      "loss": 3.904,
+      "step": 1510400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.526168414721793e-05,
+      "loss": 3.9097,
+      "step": 1510912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.525329819970741e-05,
+      "loss": 3.9119,
+      "step": 1511424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.524491225219689e-05,
+      "loss": 3.9055,
+      "step": 1511936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.523652630468637e-05,
+      "loss": 3.9185,
+      "step": 1512448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.522814035717585e-05,
+      "loss": 3.9103,
+      "step": 1512960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.521977078846906e-05,
+      "loss": 3.9098,
+      "step": 1513472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5211384840958542e-05,
+      "loss": 3.9085,
+      "step": 1513984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5202998893448022e-05,
+      "loss": 3.9113,
+      "step": 1514496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5194612945937502e-05,
+      "loss": 3.9033,
+      "step": 1515008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5186226998426982e-05,
+      "loss": 3.9171,
+      "step": 1515520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.517785742972019e-05,
+      "loss": 3.9003,
+      "step": 1516032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.516947148220967e-05,
+      "loss": 3.9218,
+      "step": 1516544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.516108553469915e-05,
+      "loss": 3.9178,
+      "step": 1517056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5152699587188634e-05,
+      "loss": 3.9133,
+      "step": 1517568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5144313639678114e-05,
+      "loss": 3.9057,
+      "step": 1518080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5135927692167594e-05,
+      "loss": 3.9115,
+      "step": 1518592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5127541744657074e-05,
+      "loss": 3.9169,
+      "step": 1519104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5119155797146554e-05,
+      "loss": 3.9161,
+      "step": 1519616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5110786228439763e-05,
+      "loss": 3.9054,
+      "step": 1520128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5102400280929243e-05,
+      "loss": 3.9195,
+      "step": 1520640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5094014333418727e-05,
+      "loss": 3.9126,
+      "step": 1521152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5085628385908206e-05,
+      "loss": 3.9197,
+      "step": 1521664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.507724243839768e-05,
+      "loss": 3.9106,
+      "step": 1522176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5068872869690896e-05,
+      "loss": 3.921,
+      "step": 1522688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5060486922180375e-05,
+      "loss": 3.9176,
+      "step": 1523200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5052100974669852e-05,
+      "loss": 3.9087,
+      "step": 1523712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5043731405963068e-05,
+      "loss": 3.9129,
+      "step": 1524224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5035345458452548e-05,
+      "loss": 3.9185,
+      "step": 1524736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5026959510942028e-05,
+      "loss": 3.9126,
+      "step": 1525248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.50185735634315e-05,
+      "loss": 3.8995,
+      "step": 1525760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.501018761592098e-05,
+      "loss": 3.9136,
+      "step": 1526272
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.003026962280273,
+      "eval_runtime": 297.0737,
+      "eval_samples_per_second": 1284.499,
+      "eval_steps_per_second": 40.142,
+      "step": 1526400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5001801668410464e-05,
+      "loss": 3.9073,
+      "step": 1526784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4993415720899944e-05,
+      "loss": 3.9028,
+      "step": 1527296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4985046152193157e-05,
+      "loss": 3.9195,
+      "step": 1527808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4976660204682637e-05,
+      "loss": 3.9187,
+      "step": 1528320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4968274257172117e-05,
+      "loss": 3.9197,
+      "step": 1528832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4959888309661593e-05,
+      "loss": 3.9108,
+      "step": 1529344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4951502362151073e-05,
+      "loss": 3.9147,
+      "step": 1529856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.494313279344429e-05,
+      "loss": 3.8969,
+      "step": 1530368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.493474684593377e-05,
+      "loss": 3.9203,
+      "step": 1530880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4926360898423246e-05,
+      "loss": 3.919,
+      "step": 1531392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4917974950912726e-05,
+      "loss": 3.9099,
+      "step": 1531904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4909589003402205e-05,
+      "loss": 3.9189,
+      "step": 1532416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4901219434695418e-05,
+      "loss": 3.9069,
+      "step": 1532928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4892833487184898e-05,
+      "loss": 3.9097,
+      "step": 1533440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4884447539674378e-05,
+      "loss": 3.9027,
+      "step": 1533952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4876061592163858e-05,
+      "loss": 3.9052,
+      "step": 1534464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4867675644653338e-05,
+      "loss": 3.9079,
+      "step": 1534976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4859289697142818e-05,
+      "loss": 3.9074,
+      "step": 1535488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4850903749632298e-05,
+      "loss": 3.9097,
+      "step": 1536000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4842517802121774e-05,
+      "loss": 3.9297,
+      "step": 1536512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.483416461221872e-05,
+      "loss": 3.9135,
+      "step": 1537024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.48257786647082e-05,
+      "loss": 3.9197,
+      "step": 1537536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.481739271719768e-05,
+      "loss": 3.9121,
+      "step": 1538048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4809023148490888e-05,
+      "loss": 3.9172,
+      "step": 1538560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.480063720098037e-05,
+      "loss": 3.9063,
+      "step": 1539072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.479225125346985e-05,
+      "loss": 3.9129,
+      "step": 1539584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.478386530595933e-05,
+      "loss": 3.9095,
+      "step": 1540096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.477547935844881e-05,
+      "loss": 3.902,
+      "step": 1540608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.476709341093829e-05,
+      "loss": 3.902,
+      "step": 1541120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.475870746342777e-05,
+      "loss": 3.9099,
+      "step": 1541632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.475033789472098e-05,
+      "loss": 3.911,
+      "step": 1542144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4741951947210464e-05,
+      "loss": 3.9197,
+      "step": 1542656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4733565999699944e-05,
+      "loss": 3.9111,
+      "step": 1543168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.472518005218942e-05,
+      "loss": 3.9143,
+      "step": 1543680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4716810483482633e-05,
+      "loss": 3.9055,
+      "step": 1544192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4708424535972113e-05,
+      "loss": 3.9198,
+      "step": 1544704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4700038588461593e-05,
+      "loss": 3.9025,
+      "step": 1545216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4691652640951073e-05,
+      "loss": 3.9111,
+      "step": 1545728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4683266693440553e-05,
+      "loss": 3.8957,
+      "step": 1546240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4674880745930033e-05,
+      "loss": 3.9097,
+      "step": 1546752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4666494798419513e-05,
+      "loss": 3.9098,
+      "step": 1547264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4658108850908993e-05,
+      "loss": 3.9196,
+      "step": 1547776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4649722903398473e-05,
+      "loss": 3.9036,
+      "step": 1548288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.464135333469168e-05,
+      "loss": 3.9102,
+      "step": 1548800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4632967387181165e-05,
+      "loss": 3.9151,
+      "step": 1549312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4624581439670645e-05,
+      "loss": 3.9141,
+      "step": 1549824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4616195492160125e-05,
+      "loss": 3.9116,
+      "step": 1550336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4607842302257066e-05,
+      "loss": 3.8991,
+      "step": 1550848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4599456354746543e-05,
+      "loss": 3.8929,
+      "step": 1551360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4591070407236026e-05,
+      "loss": 3.9224,
+      "step": 1551872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4582684459725506e-05,
+      "loss": 3.9142,
+      "step": 1552384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4574298512214986e-05,
+      "loss": 3.9068,
+      "step": 1552896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4565912564704466e-05,
+      "loss": 3.9069,
+      "step": 1553408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4557526617193946e-05,
+      "loss": 3.9057,
+      "step": 1553920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4549140669683426e-05,
+      "loss": 3.8887,
+      "step": 1554432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4540754722172903e-05,
+      "loss": 3.9067,
+      "step": 1554944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4532368774662383e-05,
+      "loss": 3.8953,
+      "step": 1555456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4523982827151866e-05,
+      "loss": 3.9081,
+      "step": 1555968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4515596879641346e-05,
+      "loss": 3.9155,
+      "step": 1556480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4507210932130826e-05,
+      "loss": 3.9024,
+      "step": 1556992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4498841363424035e-05,
+      "loss": 3.8936,
+      "step": 1557504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4490455415913515e-05,
+      "loss": 3.9129,
+      "step": 1558016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4482069468402995e-05,
+      "loss": 3.891,
+      "step": 1558528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4473683520892475e-05,
+      "loss": 3.901,
+      "step": 1559040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4465313952185687e-05,
+      "loss": 3.9087,
+      "step": 1559552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4456928004675167e-05,
+      "loss": 3.9086,
+      "step": 1560064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4448542057164647e-05,
+      "loss": 3.8975,
+      "step": 1560576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4440156109654127e-05,
+      "loss": 3.8974,
+      "step": 1561088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4431786540947336e-05,
+      "loss": 3.8973,
+      "step": 1561600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.442340059343682e-05,
+      "loss": 3.9098,
+      "step": 1562112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.44150146459263e-05,
+      "loss": 3.909,
+      "step": 1562624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.440662869841578e-05,
+      "loss": 3.9062,
+      "step": 1563136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.439825912970899e-05,
+      "loss": 3.9122,
+      "step": 1563648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.438987318219847e-05,
+      "loss": 3.9133,
+      "step": 1564160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.438148723468795e-05,
+      "loss": 3.9049,
+      "step": 1564672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.437310128717743e-05,
+      "loss": 3.898,
+      "step": 1565184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.436471533966691e-05,
+      "loss": 3.9153,
+      "step": 1565696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.435632939215639e-05,
+      "loss": 3.909,
+      "step": 1566208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.43479598234496e-05,
+      "loss": 3.9027,
+      "step": 1566720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.433957387593908e-05,
+      "loss": 3.9019,
+      "step": 1567232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4331187928428558e-05,
+      "loss": 3.9108,
+      "step": 1567744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4322818359721773e-05,
+      "loss": 3.9189,
+      "step": 1568256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4314432412211253e-05,
+      "loss": 3.9012,
+      "step": 1568768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.430604646470073e-05,
+      "loss": 3.89,
+      "step": 1569280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.429766051719021e-05,
+      "loss": 3.8996,
+      "step": 1569792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.428927456967969e-05,
+      "loss": 3.8986,
+      "step": 1570304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.428088862216917e-05,
+      "loss": 3.9079,
+      "step": 1570816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.427250267465865e-05,
+      "loss": 3.9079,
+      "step": 1571328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.426411672714813e-05,
+      "loss": 3.8951,
+      "step": 1571840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4255747158441342e-05,
+      "loss": 3.911,
+      "step": 1572352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4247361210930822e-05,
+      "loss": 3.9012,
+      "step": 1572864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4238975263420302e-05,
+      "loss": 3.9034,
+      "step": 1573376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4230589315909782e-05,
+      "loss": 3.8992,
+      "step": 1573888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4222203368399262e-05,
+      "loss": 3.9072,
+      "step": 1574400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4213817420888742e-05,
+      "loss": 3.9179,
+      "step": 1574912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4205431473378222e-05,
+      "loss": 3.9068,
+      "step": 1575424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4197045525867702e-05,
+      "loss": 3.8861,
+      "step": 1575936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.418867595716091e-05,
+      "loss": 3.9057,
+      "step": 1576448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.418029000965039e-05,
+      "loss": 3.9036,
+      "step": 1576960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.417190406213987e-05,
+      "loss": 3.9111,
+      "step": 1577472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4163534493433083e-05,
+      "loss": 3.9132,
+      "step": 1577984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4155148545922563e-05,
+      "loss": 3.9027,
+      "step": 1578496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4146762598412043e-05,
+      "loss": 3.8962,
+      "step": 1579008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4138376650901523e-05,
+      "loss": 3.9092,
+      "step": 1579520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4130007082194732e-05,
+      "loss": 3.8879,
+      "step": 1580032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4121637513487945e-05,
+      "loss": 3.904,
+      "step": 1580544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4113251565977428e-05,
+      "loss": 3.8982,
+      "step": 1581056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4104865618466908e-05,
+      "loss": 3.9084,
+      "step": 1581568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4096479670956385e-05,
+      "loss": 3.8918,
+      "step": 1582080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4088093723445865e-05,
+      "loss": 3.9099,
+      "step": 1582592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4079707775935345e-05,
+      "loss": 3.895,
+      "step": 1583104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4071321828424825e-05,
+      "loss": 3.8978,
+      "step": 1583616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4062935880914304e-05,
+      "loss": 3.9097,
+      "step": 1584128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4054566312207517e-05,
+      "loss": 3.9013,
+      "step": 1584640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4046180364696997e-05,
+      "loss": 3.9195,
+      "step": 1585152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4037794417186477e-05,
+      "loss": 3.911,
+      "step": 1585664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4029408469675957e-05,
+      "loss": 3.9022,
+      "step": 1586176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4021022522165437e-05,
+      "loss": 3.906,
+      "step": 1586688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4012652953458646e-05,
+      "loss": 3.8968,
+      "step": 1587200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.400426700594813e-05,
+      "loss": 3.9058,
+      "step": 1587712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.399588105843761e-05,
+      "loss": 3.8953,
+      "step": 1588224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.398749511092709e-05,
+      "loss": 3.9124,
+      "step": 1588736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3979109163416566e-05,
+      "loss": 3.8974,
+      "step": 1589248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3970723215906046e-05,
+      "loss": 3.8997,
+      "step": 1589760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3962353647199258e-05,
+      "loss": 3.8993,
+      "step": 1590272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.395398407849247e-05,
+      "loss": 3.9062,
+      "step": 1590784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.394559813098195e-05,
+      "loss": 3.8957,
+      "step": 1591296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.393721218347143e-05,
+      "loss": 3.9079,
+      "step": 1591808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.392882623596091e-05,
+      "loss": 3.8928,
+      "step": 1592320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3920440288450387e-05,
+      "loss": 3.9136,
+      "step": 1592832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3912054340939867e-05,
+      "loss": 3.9089,
+      "step": 1593344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.390366839342935e-05,
+      "loss": 3.9061,
+      "step": 1593856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.389528244591883e-05,
+      "loss": 3.8999,
+      "step": 1594368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.388689649840831e-05,
+      "loss": 3.9017,
+      "step": 1594880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.387852692970152e-05,
+      "loss": 3.9136,
+      "step": 1595392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3870140982191e-05,
+      "loss": 3.9058,
+      "step": 1595904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.386175503468048e-05,
+      "loss": 3.897,
+      "step": 1596416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3853385465973692e-05,
+      "loss": 3.9113,
+      "step": 1596928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3844999518463172e-05,
+      "loss": 3.9071,
+      "step": 1597440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.383661357095265e-05,
+      "loss": 3.9146,
+      "step": 1597952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.382822762344213e-05,
+      "loss": 3.9027,
+      "step": 1598464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.381984167593161e-05,
+      "loss": 3.9123,
+      "step": 1598976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.381145572842109e-05,
+      "loss": 3.909,
+      "step": 1599488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.38030861597143e-05,
+      "loss": 3.9019,
+      "step": 1600000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3794700212203784e-05,
+      "loss": 3.9075,
+      "step": 1600512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3786330643496993e-05,
+      "loss": 3.9084,
+      "step": 1601024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3777944695986473e-05,
+      "loss": 3.9052,
+      "step": 1601536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3769558748475953e-05,
+      "loss": 3.8915,
+      "step": 1602048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3761172800965433e-05,
+      "loss": 3.9056,
+      "step": 1602560
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.000296592712402,
+      "eval_runtime": 297.3335,
+      "eval_samples_per_second": 1283.377,
+      "eval_steps_per_second": 40.106,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3752786853454913e-05,
+      "loss": 3.8971,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3744400905944393e-05,
+      "loss": 3.8954,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3736014958433873e-05,
+      "loss": 3.9119,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3727629010923353e-05,
+      "loss": 3.9127,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3719243063412833e-05,
+      "loss": 3.9098,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3710857115902313e-05,
+      "loss": 3.904,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3702471168391793e-05,
+      "loss": 3.9054,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3694085220881273e-05,
+      "loss": 3.892,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3685699273370753e-05,
+      "loss": 3.9096,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3677313325860233e-05,
+      "loss": 3.9132,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3668927378349712e-05,
+      "loss": 3.9028,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3660541430839192e-05,
+      "loss": 3.9103,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.36521718621324e-05,
+      "loss": 3.9024,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.364378591462188e-05,
+      "loss": 3.9036,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.363539996711136e-05,
+      "loss": 3.8947,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3627014019600845e-05,
+      "loss": 3.8969,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3618628072090325e-05,
+      "loss": 3.8973,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3610242124579805e-05,
+      "loss": 3.8994,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3601856177069285e-05,
+      "loss": 3.9045,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.359347022955876e-05,
+      "loss": 3.9169,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3585100660851974e-05,
+      "loss": 3.9125,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3576731092145186e-05,
+      "loss": 3.9114,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3568345144634666e-05,
+      "loss": 3.9022,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3559959197124146e-05,
+      "loss": 3.9125,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3551573249613626e-05,
+      "loss": 3.8969,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3543187302103106e-05,
+      "loss": 3.9095,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3534801354592583e-05,
+      "loss": 3.9007,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3526415407082063e-05,
+      "loss": 3.8922,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3518029459571546e-05,
+      "loss": 3.8993,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3509643512061026e-05,
+      "loss": 3.8976,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3501257564550506e-05,
+      "loss": 3.9033,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3492887995843715e-05,
+      "loss": 3.9087,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3484502048333195e-05,
+      "loss": 3.9089,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3476116100822675e-05,
+      "loss": 3.9075,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3467746532115887e-05,
+      "loss": 3.8978,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3459360584605367e-05,
+      "loss": 3.9104,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3450974637094847e-05,
+      "loss": 3.8979,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3442588689584327e-05,
+      "loss": 3.8959,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3434202742073807e-05,
+      "loss": 3.8925,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3425816794563287e-05,
+      "loss": 3.8989,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3417430847052767e-05,
+      "loss": 3.9052,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.340906127834598e-05,
+      "loss": 3.9093,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.340067533083546e-05,
+      "loss": 3.8981,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3392289383324936e-05,
+      "loss": 3.9006,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3383903435814416e-05,
+      "loss": 3.911,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3375517488303896e-05,
+      "loss": 3.905,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3367131540793376e-05,
+      "loss": 3.9076,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.335876197208659e-05,
+      "loss": 3.8938,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.335037602457607e-05,
+      "loss": 3.8817,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3341990077065548e-05,
+      "loss": 3.9174,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3333604129555028e-05,
+      "loss": 3.9073,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3325218182044508e-05,
+      "loss": 3.9014,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3316832234533988e-05,
+      "loss": 3.9008,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3308446287023468e-05,
+      "loss": 3.8942,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3300060339512948e-05,
+      "loss": 3.8866,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3291674392002428e-05,
+      "loss": 3.8981,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.328330482329564e-05,
+      "loss": 3.8895,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.327493525458885e-05,
+      "loss": 3.8993,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.326654930707833e-05,
+      "loss": 3.9091,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.325816335956781e-05,
+      "loss": 3.8959,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3249777412057293e-05,
+      "loss": 3.887,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.324139146454677e-05,
+      "loss": 3.9063,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.323300551703625e-05,
+      "loss": 3.8852,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.322461956952573e-05,
+      "loss": 3.8908,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.321623362201521e-05,
+      "loss": 3.9033,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3207864053308422e-05,
+      "loss": 3.9023,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3199478105797902e-05,
+      "loss": 3.8907,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.319109215828738e-05,
+      "loss": 3.8905,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.318270621077686e-05,
+      "loss": 3.8905,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.317432026326634e-05,
+      "loss": 3.8977,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.316593431575582e-05,
+      "loss": 3.9045,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3157548368245298e-05,
+      "loss": 3.8953,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3149162420734778e-05,
+      "loss": 3.9039,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.314077647322426e-05,
+      "loss": 3.9064,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3132406904517474e-05,
+      "loss": 3.9045,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.312402095700695e-05,
+      "loss": 3.8878,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3115651388300163e-05,
+      "loss": 3.9057,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3107265440789643e-05,
+      "loss": 3.9052,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3098879493279123e-05,
+      "loss": 3.8908,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3090493545768603e-05,
+      "loss": 3.9005,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3082123977061815e-05,
+      "loss": 3.8999,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3073754408355024e-05,
+      "loss": 3.91,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3065368460844504e-05,
+      "loss": 3.8986,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3056982513333984e-05,
+      "loss": 3.8776,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3048596565823464e-05,
+      "loss": 3.8945,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3040210618312948e-05,
+      "loss": 3.8924,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3031824670802424e-05,
+      "loss": 3.899,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3023438723291904e-05,
+      "loss": 3.9011,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3015052775781384e-05,
+      "loss": 3.8903,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3006666828270864e-05,
+      "loss": 3.8991,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2998280880760344e-05,
+      "loss": 3.8973,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2989894933249824e-05,
+      "loss": 3.8897,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2981508985739304e-05,
+      "loss": 3.8948,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2973139417032516e-05,
+      "loss": 3.9026,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2964753469521996e-05,
+      "loss": 3.9099,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2956367522011476e-05,
+      "loss": 3.8992,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2947981574500953e-05,
+      "loss": 3.8807,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2939595626990433e-05,
+      "loss": 3.8975,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2931209679479916e-05,
+      "loss": 3.8959,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2922823731969396e-05,
+      "loss": 3.9043,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2914437784458876e-05,
+      "loss": 3.9064,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2906068215752085e-05,
+      "loss": 3.8891,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2897682268241565e-05,
+      "loss": 3.8903,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2889296320731045e-05,
+      "loss": 3.9027,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2880910373220525e-05,
+      "loss": 3.8803,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2872540804513738e-05,
+      "loss": 3.8984,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2864154857003217e-05,
+      "loss": 3.8858,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2855768909492697e-05,
+      "loss": 3.9018,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2847382961982177e-05,
+      "loss": 3.886,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2839013393275386e-05,
+      "loss": 3.9029,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2830643824568602e-05,
+      "loss": 3.8942,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.282225787705808e-05,
+      "loss": 3.8864,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.281387192954756e-05,
+      "loss": 3.9042,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.280548598203704e-05,
+      "loss": 3.8914,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.279710003452652e-05,
+      "loss": 3.9112,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2788714087016e-05,
+      "loss": 3.908,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.278032813950548e-05,
+      "loss": 3.8928,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.277194219199496e-05,
+      "loss": 3.9008,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.27635890020919e-05,
+      "loss": 3.8874,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.275520305458138e-05,
+      "loss": 3.8958,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274681710707086e-05,
+      "loss": 3.8935,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.273843115956034e-05,
+      "loss": 3.9025,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2730045212049823e-05,
+      "loss": 3.8944,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2721659264539303e-05,
+      "loss": 3.8906,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2713273317028783e-05,
+      "loss": 3.896,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2704903748321993e-05,
+      "loss": 3.8966,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2696517800811472e-05,
+      "loss": 3.8864,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2688131853300952e-05,
+      "loss": 3.9038,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2679762284594165e-05,
+      "loss": 3.8841,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2671376337083645e-05,
+      "loss": 3.9122,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2662990389573125e-05,
+      "loss": 3.8936,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2654604442062605e-05,
+      "loss": 3.9032,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.264621849455208e-05,
+      "loss": 3.8903,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.263783254704156e-05,
+      "loss": 3.8991,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262944659953104e-05,
+      "loss": 3.9046,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2621060652020525e-05,
+      "loss": 3.8968,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2612691083313734e-05,
+      "loss": 3.8924,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2604305135803214e-05,
+      "loss": 3.9059,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2595919188292694e-05,
+      "loss": 3.8999,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2587533240782174e-05,
+      "loss": 3.9066,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2579147293271654e-05,
+      "loss": 3.8922,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2570761345761133e-05,
+      "loss": 3.9046,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2562375398250613e-05,
+      "loss": 3.9003,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2554022208347555e-05,
+      "loss": 3.897,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2545636260837035e-05,
+      "loss": 3.9014,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2537250313326515e-05,
+      "loss": 3.8993,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2528864365815995e-05,
+      "loss": 3.9054,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2520478418305478e-05,
+      "loss": 3.8773,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2512092470794958e-05,
+      "loss": 3.9037,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9974029064178467,
+      "eval_runtime": 305.8486,
+      "eval_samples_per_second": 1247.647,
+      "eval_steps_per_second": 38.99,
+      "step": 1679040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2503706523284438e-05,
+      "loss": 3.895,
+      "step": 1679360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2495320575773915e-05,
+      "loss": 3.8887,
+      "step": 1679872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2486934628263395e-05,
+      "loss": 3.9019,
+      "step": 1680384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2478548680752875e-05,
+      "loss": 3.9078,
+      "step": 1680896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2470162733242355e-05,
+      "loss": 3.9013,
+      "step": 1681408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2461776785731835e-05,
+      "loss": 3.8927,
+      "step": 1681920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2453407217025047e-05,
+      "loss": 3.8979,
+      "step": 1682432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2445021269514527e-05,
+      "loss": 3.8859,
+      "step": 1682944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2436635322004007e-05,
+      "loss": 3.9001,
+      "step": 1683456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2428249374493487e-05,
+      "loss": 3.904,
+      "step": 1683968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2419863426982967e-05,
+      "loss": 3.9018,
+      "step": 1684480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.241149385827618e-05,
+      "loss": 3.9012,
+      "step": 1684992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.240312428956939e-05,
+      "loss": 3.8983,
+      "step": 1685504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.239473834205887e-05,
+      "loss": 3.8926,
+      "step": 1686016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.238635239454835e-05,
+      "loss": 3.886,
+      "step": 1686528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.237796644703783e-05,
+      "loss": 3.8858,
+      "step": 1687040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2369580499527308e-05,
+      "loss": 3.8951,
+      "step": 1687552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2361194552016788e-05,
+      "loss": 3.8917,
+      "step": 1688064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2352808604506268e-05,
+      "loss": 3.8971,
+      "step": 1688576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2344422656995748e-05,
+      "loss": 3.9098,
+      "step": 1689088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2336036709485228e-05,
+      "loss": 3.91,
+      "step": 1689600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.232766714077844e-05,
+      "loss": 3.901,
+      "step": 1690112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2319281193267917e-05,
+      "loss": 3.8955,
+      "step": 1690624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2310911624561133e-05,
+      "loss": 3.9043,
+      "step": 1691136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2302525677050613e-05,
+      "loss": 3.8909,
+      "step": 1691648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.229413972954009e-05,
+      "loss": 3.9048,
+      "step": 1692160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.228575378202957e-05,
+      "loss": 3.894,
+      "step": 1692672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.227736783451905e-05,
+      "loss": 3.8822,
+      "step": 1693184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2268998265812262e-05,
+      "loss": 3.8951,
+      "step": 1693696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2260612318301742e-05,
+      "loss": 3.8909,
+      "step": 1694208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2252226370791222e-05,
+      "loss": 3.8936,
+      "step": 1694720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2243840423280702e-05,
+      "loss": 3.9035,
+      "step": 1695232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2235454475770182e-05,
+      "loss": 3.8997,
+      "step": 1695744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.222708490706339e-05,
+      "loss": 3.9007,
+      "step": 1696256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2218715338356603e-05,
+      "loss": 3.8893,
+      "step": 1696768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2210329390846087e-05,
+      "loss": 3.9097,
+      "step": 1697280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2201943443335563e-05,
+      "loss": 3.89,
+      "step": 1697792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2193557495825043e-05,
+      "loss": 3.8852,
+      "step": 1698304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2185171548314523e-05,
+      "loss": 3.8902,
+      "step": 1698816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2176785600804003e-05,
+      "loss": 3.8855,
+      "step": 1699328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2168416032097216e-05,
+      "loss": 3.8985,
+      "step": 1699840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2160030084586696e-05,
+      "loss": 3.9053,
+      "step": 1700352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2151644137076176e-05,
+      "loss": 3.8887,
+      "step": 1700864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2143258189565655e-05,
+      "loss": 3.8942,
+      "step": 1701376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2134872242055135e-05,
+      "loss": 3.9033,
+      "step": 1701888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2126486294544615e-05,
+      "loss": 3.9015,
+      "step": 1702400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2118100347034095e-05,
+      "loss": 3.9011,
+      "step": 1702912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2109730778327304e-05,
+      "loss": 3.889,
+      "step": 1703424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2101344830816788e-05,
+      "loss": 3.8677,
+      "step": 1703936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2092958883306268e-05,
+      "loss": 3.9104,
+      "step": 1704448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2084572935795744e-05,
+      "loss": 3.9003,
+      "step": 1704960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2076186988285224e-05,
+      "loss": 3.8967,
+      "step": 1705472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2067801040774704e-05,
+      "loss": 3.8913,
+      "step": 1705984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2059415093264184e-05,
+      "loss": 3.887,
+      "step": 1706496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2051029145753664e-05,
+      "loss": 3.8809,
+      "step": 1707008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2042643198243147e-05,
+      "loss": 3.8871,
+      "step": 1707520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2034257250732627e-05,
+      "loss": 3.8843,
+      "step": 1708032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2025871303222104e-05,
+      "loss": 3.8906,
+      "step": 1708544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2017501734515316e-05,
+      "loss": 3.9029,
+      "step": 1709056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2009115787004796e-05,
+      "loss": 3.886,
+      "step": 1709568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2000729839494276e-05,
+      "loss": 3.8815,
+      "step": 1710080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1992343891983756e-05,
+      "loss": 3.8992,
+      "step": 1710592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1983957944473236e-05,
+      "loss": 3.8831,
+      "step": 1711104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1975571996962716e-05,
+      "loss": 3.8841,
+      "step": 1711616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1967186049452196e-05,
+      "loss": 3.8924,
+      "step": 1712128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1958800101941676e-05,
+      "loss": 3.8949,
+      "step": 1712640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1950430533234885e-05,
+      "loss": 3.8831,
+      "step": 1713152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1942044585724365e-05,
+      "loss": 3.881,
+      "step": 1713664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.193365863821385e-05,
+      "loss": 3.8814,
+      "step": 1714176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.192527269070333e-05,
+      "loss": 3.8959,
+      "step": 1714688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1916903121996538e-05,
+      "loss": 3.8939,
+      "step": 1715200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1908517174486018e-05,
+      "loss": 3.8923,
+      "step": 1715712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1900131226975498e-05,
+      "loss": 3.8931,
+      "step": 1716224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.189176165826871e-05,
+      "loss": 3.901,
+      "step": 1716736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1883392089561922e-05,
+      "loss": 3.8995,
+      "step": 1717248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.18750061420514e-05,
+      "loss": 3.8838,
+      "step": 1717760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.186662019454088e-05,
+      "loss": 3.8958,
+      "step": 1718272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.185823424703036e-05,
+      "loss": 3.8992,
+      "step": 1718784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.184984829951984e-05,
+      "loss": 3.8816,
+      "step": 1719296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.184146235200932e-05,
+      "loss": 3.8943,
+      "step": 1719808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1833076404498802e-05,
+      "loss": 3.8932,
+      "step": 1720320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1824690456988282e-05,
+      "loss": 3.9016,
+      "step": 1720832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.181630450947776e-05,
+      "loss": 3.8931,
+      "step": 1721344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.180791856196724e-05,
+      "loss": 3.8731,
+      "step": 1721856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.179953261445672e-05,
+      "loss": 3.8875,
+      "step": 1722368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1791163045749928e-05,
+      "loss": 3.8843,
+      "step": 1722880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.178277709823941e-05,
+      "loss": 3.8951,
+      "step": 1723392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.177439115072889e-05,
+      "loss": 3.8902,
+      "step": 1723904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.176600520321837e-05,
+      "loss": 3.8854,
+      "step": 1724416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.175763563451158e-05,
+      "loss": 3.8908,
+      "step": 1724928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.174924968700106e-05,
+      "loss": 3.8966,
+      "step": 1725440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.174086373949054e-05,
+      "loss": 3.8811,
+      "step": 1725952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.173247779198002e-05,
+      "loss": 3.8852,
+      "step": 1726464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1724108223273232e-05,
+      "loss": 3.8994,
+      "step": 1726976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1715722275762712e-05,
+      "loss": 3.9005,
+      "step": 1727488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1707336328252192e-05,
+      "loss": 3.8915,
+      "step": 1728000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1698950380741672e-05,
+      "loss": 3.8767,
+      "step": 1728512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1690564433231152e-05,
+      "loss": 3.8875,
+      "step": 1729024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1682178485720632e-05,
+      "loss": 3.8845,
+      "step": 1729536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1673792538210112e-05,
+      "loss": 3.8983,
+      "step": 1730048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1665422969503325e-05,
+      "loss": 3.8996,
+      "step": 1730560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1657037021992805e-05,
+      "loss": 3.8856,
+      "step": 1731072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1648651074482285e-05,
+      "loss": 3.8833,
+      "step": 1731584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.164026512697176e-05,
+      "loss": 3.8927,
+      "step": 1732096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.163187917946124e-05,
+      "loss": 3.875,
+      "step": 1732608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.162349323195072e-05,
+      "loss": 3.8933,
+      "step": 1733120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1615123663243937e-05,
+      "loss": 3.8757,
+      "step": 1733632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1606737715733414e-05,
+      "loss": 3.8961,
+      "step": 1734144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1598351768222893e-05,
+      "loss": 3.8832,
+      "step": 1734656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1589982199516106e-05,
+      "loss": 3.891,
+      "step": 1735168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1581596252005586e-05,
+      "loss": 3.8867,
+      "step": 1735680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1573210304495066e-05,
+      "loss": 3.885,
+      "step": 1736192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1564824356984546e-05,
+      "loss": 3.8936,
+      "step": 1736704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1556438409474026e-05,
+      "loss": 3.8821,
+      "step": 1737216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1548052461963506e-05,
+      "loss": 3.9046,
+      "step": 1737728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1539666514452986e-05,
+      "loss": 3.8977,
+      "step": 1738240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1531280566942466e-05,
+      "loss": 3.8881,
+      "step": 1738752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1522910998235675e-05,
+      "loss": 3.8938,
+      "step": 1739264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1514525050725158e-05,
+      "loss": 3.8808,
+      "step": 1739776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1506139103214638e-05,
+      "loss": 3.8893,
+      "step": 1740288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1497753155704118e-05,
+      "loss": 3.8898,
+      "step": 1740800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1489367208193595e-05,
+      "loss": 3.8918,
+      "step": 1741312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1480981260683075e-05,
+      "loss": 3.8879,
+      "step": 1741824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1472595313172554e-05,
+      "loss": 3.8862,
+      "step": 1742336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1464209365662034e-05,
+      "loss": 3.8909,
+      "step": 1742848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1455823418151518e-05,
+      "loss": 3.8879,
+      "step": 1743360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1447453849444727e-05,
+      "loss": 3.8836,
+      "step": 1743872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1439067901934207e-05,
+      "loss": 3.8934,
+      "step": 1744384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1430681954423687e-05,
+      "loss": 3.8764,
+      "step": 1744896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1422296006913167e-05,
+      "loss": 3.9055,
+      "step": 1745408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.141392643820638e-05,
+      "loss": 3.892,
+      "step": 1745920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.140554049069586e-05,
+      "loss": 3.8959,
+      "step": 1746432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.139715454318534e-05,
+      "loss": 3.8796,
+      "step": 1746944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.138876859567482e-05,
+      "loss": 3.8943,
+      "step": 1747456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1380399026968028e-05,
+      "loss": 3.8952,
+      "step": 1747968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1372013079457508e-05,
+      "loss": 3.8933,
+      "step": 1748480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1363627131946988e-05,
+      "loss": 3.8833,
+      "step": 1748992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1355241184436468e-05,
+      "loss": 3.8977,
+      "step": 1749504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1346855236925948e-05,
+      "loss": 3.8922,
+      "step": 1750016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1338469289415428e-05,
+      "loss": 3.9018,
+      "step": 1750528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1330083341904908e-05,
+      "loss": 3.8841,
+      "step": 1751040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1321697394394388e-05,
+      "loss": 3.9003,
+      "step": 1751552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1313311446883868e-05,
+      "loss": 3.8939,
+      "step": 1752064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1304958256980813e-05,
+      "loss": 3.8894,
+      "step": 1752576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1296572309470293e-05,
+      "loss": 3.8916,
+      "step": 1753088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1288186361959773e-05,
+      "loss": 3.8933,
+      "step": 1753600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.127980041444925e-05,
+      "loss": 3.8946,
+      "step": 1754112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.127141446693873e-05,
+      "loss": 3.8766,
+      "step": 1754624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.126302851942821e-05,
+      "loss": 3.8981,
+      "step": 1755136
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.994722366333008,
+      "eval_runtime": 307.1394,
+      "eval_samples_per_second": 1242.403,
+      "eval_steps_per_second": 38.826,
+      "step": 1755360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.125464257191769e-05,
+      "loss": 3.8903,
+      "step": 1755648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1246256624407173e-05,
+      "loss": 3.882,
+      "step": 1756160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1237870676896652e-05,
+      "loss": 3.8944,
+      "step": 1756672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.122948472938613e-05,
+      "loss": 3.9001,
+      "step": 1757184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.122109878187561e-05,
+      "loss": 3.8963,
+      "step": 1757696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.121271283436509e-05,
+      "loss": 3.8888,
+      "step": 1758208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.120432688685457e-05,
+      "loss": 3.895,
+      "step": 1758720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.119594093934405e-05,
+      "loss": 3.8815,
+      "step": 1759232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.118755499183353e-05,
+      "loss": 3.895,
+      "step": 1759744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1179169044323012e-05,
+      "loss": 3.8978,
+      "step": 1760256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.117078309681249e-05,
+      "loss": 3.8932,
+      "step": 1760768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.116239714930197e-05,
+      "loss": 3.8964,
+      "step": 1761280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.115402758059518e-05,
+      "loss": 3.8883,
+      "step": 1761792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1145641633084658e-05,
+      "loss": 3.8852,
+      "step": 1762304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1137255685574138e-05,
+      "loss": 3.8861,
+      "step": 1762816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.112886973806362e-05,
+      "loss": 3.8769,
+      "step": 1763328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.11204837905531e-05,
+      "loss": 3.8884,
+      "step": 1763840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.111209784304258e-05,
+      "loss": 3.8842,
+      "step": 1764352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.110371189553206e-05,
+      "loss": 3.8886,
+      "step": 1764864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.109532594802154e-05,
+      "loss": 3.9019,
+      "step": 1765376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1086940000511017e-05,
+      "loss": 3.9037,
+      "step": 1765888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1078554053000497e-05,
+      "loss": 3.899,
+      "step": 1766400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.107016810548998e-05,
+      "loss": 3.8879,
+      "step": 1766912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1061798536783193e-05,
+      "loss": 3.8963,
+      "step": 1767424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.105341258927267e-05,
+      "loss": 3.8865,
+      "step": 1767936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.104502664176215e-05,
+      "loss": 3.8964,
+      "step": 1768448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.103664069425163e-05,
+      "loss": 3.8891,
+      "step": 1768960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1028271125544842e-05,
+      "loss": 3.8778,
+      "step": 1769472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1019885178034322e-05,
+      "loss": 3.8911,
+      "step": 1769984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1011499230523802e-05,
+      "loss": 3.8803,
+      "step": 1770496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1003113283013282e-05,
+      "loss": 3.8884,
+      "step": 1771008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0994727335502762e-05,
+      "loss": 3.8975,
+      "step": 1771520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0986341387992242e-05,
+      "loss": 3.8918,
+      "step": 1772032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0977955440481722e-05,
+      "loss": 3.8931,
+      "step": 1772544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.09695694929712e-05,
+      "loss": 3.8824,
+      "step": 1773056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0961199924264414e-05,
+      "loss": 3.9054,
+      "step": 1773568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0952813976753894e-05,
+      "loss": 3.8869,
+      "step": 1774080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0944428029243374e-05,
+      "loss": 3.878,
+      "step": 1774592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.093604208173285e-05,
+      "loss": 3.8876,
+      "step": 1775104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.092765613422233e-05,
+      "loss": 3.8783,
+      "step": 1775616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.091927018671181e-05,
+      "loss": 3.8898,
+      "step": 1776128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.091088423920129e-05,
+      "loss": 3.8959,
+      "step": 1776640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0902514670494503e-05,
+      "loss": 3.8856,
+      "step": 1777152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0894128722983983e-05,
+      "loss": 3.886,
+      "step": 1777664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0885742775473463e-05,
+      "loss": 3.899,
+      "step": 1778176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0877356827962943e-05,
+      "loss": 3.8938,
+      "step": 1778688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0868970880452423e-05,
+      "loss": 3.897,
+      "step": 1779200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0860584932941903e-05,
+      "loss": 3.8809,
+      "step": 1779712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0852198985431383e-05,
+      "loss": 3.8591,
+      "step": 1780224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0843829416724595e-05,
+      "loss": 3.9086,
+      "step": 1780736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0835443469214075e-05,
+      "loss": 3.8902,
+      "step": 1781248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0827057521703555e-05,
+      "loss": 3.894,
+      "step": 1781760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0818671574193032e-05,
+      "loss": 3.8873,
+      "step": 1782272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0810285626682512e-05,
+      "loss": 3.878,
+      "step": 1782784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0801899679171992e-05,
+      "loss": 3.8775,
+      "step": 1783296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0793530110465204e-05,
+      "loss": 3.8819,
+      "step": 1783808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0785144162954684e-05,
+      "loss": 3.8766,
+      "step": 1784320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0776758215444164e-05,
+      "loss": 3.8829,
+      "step": 1784832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0768372267933644e-05,
+      "loss": 3.9006,
+      "step": 1785344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0759986320423124e-05,
+      "loss": 3.8775,
+      "step": 1785856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0751600372912604e-05,
+      "loss": 3.8729,
+      "step": 1786368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0743214425402084e-05,
+      "loss": 3.8974,
+      "step": 1786880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0734844856695296e-05,
+      "loss": 3.8727,
+      "step": 1787392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0726458909184776e-05,
+      "loss": 3.8777,
+      "step": 1787904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0718072961674256e-05,
+      "loss": 3.8871,
+      "step": 1788416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0709703392967465e-05,
+      "loss": 3.8894,
+      "step": 1788928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0701317445456945e-05,
+      "loss": 3.8754,
+      "step": 1789440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.069293149794643e-05,
+      "loss": 3.8829,
+      "step": 1789952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.068454555043591e-05,
+      "loss": 3.8702,
+      "step": 1790464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0676159602925385e-05,
+      "loss": 3.89,
+      "step": 1790976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0667773655414865e-05,
+      "loss": 3.8875,
+      "step": 1791488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0659387707904345e-05,
+      "loss": 3.8838,
+      "step": 1792000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0651001760393825e-05,
+      "loss": 3.8899,
+      "step": 1792512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0642615812883305e-05,
+      "loss": 3.8899,
+      "step": 1793024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0634246244176518e-05,
+      "loss": 3.898,
+      "step": 1793536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0625860296665998e-05,
+      "loss": 3.8746,
+      "step": 1794048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0617474349155478e-05,
+      "loss": 3.8903,
+      "step": 1794560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0609088401644957e-05,
+      "loss": 3.8948,
+      "step": 1795072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0600718832938167e-05,
+      "loss": 3.874,
+      "step": 1795584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0592332885427647e-05,
+      "loss": 3.8923,
+      "step": 1796096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.058394693791713e-05,
+      "loss": 3.8852,
+      "step": 1796608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.057556099040661e-05,
+      "loss": 3.8965,
+      "step": 1797120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.056717504289609e-05,
+      "loss": 3.8918,
+      "step": 1797632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0558789095385566e-05,
+      "loss": 3.8722,
+      "step": 1798144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0550403147875046e-05,
+      "loss": 3.8733,
+      "step": 1798656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0542017200364526e-05,
+      "loss": 3.882,
+      "step": 1799168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0533631252854006e-05,
+      "loss": 3.8912,
+      "step": 1799680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.052526168414722e-05,
+      "loss": 3.8821,
+      "step": 1800192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.05168757366367e-05,
+      "loss": 3.8798,
+      "step": 1800704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.050850616792991e-05,
+      "loss": 3.8852,
+      "step": 1801216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0500120220419388e-05,
+      "loss": 3.8874,
+      "step": 1801728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0491734272908868e-05,
+      "loss": 3.8787,
+      "step": 1802240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.048334832539835e-05,
+      "loss": 3.8804,
+      "step": 1802752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0474978756691564e-05,
+      "loss": 3.8906,
+      "step": 1803264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.046659280918104e-05,
+      "loss": 3.8917,
+      "step": 1803776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0458223240474253e-05,
+      "loss": 3.8884,
+      "step": 1804288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0449837292963733e-05,
+      "loss": 3.8731,
+      "step": 1804800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0441451345453212e-05,
+      "loss": 3.8848,
+      "step": 1805312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0433065397942692e-05,
+      "loss": 3.878,
+      "step": 1805824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0424679450432172e-05,
+      "loss": 3.8936,
+      "step": 1806336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0416293502921652e-05,
+      "loss": 3.8892,
+      "step": 1806848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0407907555411132e-05,
+      "loss": 3.8842,
+      "step": 1807360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0399521607900612e-05,
+      "loss": 3.8799,
+      "step": 1807872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.039115203919382e-05,
+      "loss": 3.8812,
+      "step": 1808384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.03827660916833e-05,
+      "loss": 3.8742,
+      "step": 1808896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0374380144172785e-05,
+      "loss": 3.8834,
+      "step": 1809408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0365994196662265e-05,
+      "loss": 3.8712,
+      "step": 1809920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0357608249151745e-05,
+      "loss": 3.8906,
+      "step": 1810432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.034922230164122e-05,
+      "loss": 3.8786,
+      "step": 1810944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.03408363541307e-05,
+      "loss": 3.8829,
+      "step": 1811456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0332466785423914e-05,
+      "loss": 3.8799,
+      "step": 1811968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0324080837913394e-05,
+      "loss": 3.8776,
+      "step": 1812480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0315694890402873e-05,
+      "loss": 3.8905,
+      "step": 1812992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0307325321696086e-05,
+      "loss": 3.8787,
+      "step": 1813504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0298939374185566e-05,
+      "loss": 3.9002,
+      "step": 1814016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0290553426675042e-05,
+      "loss": 3.8926,
+      "step": 1814528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0282167479164522e-05,
+      "loss": 3.8835,
+      "step": 1815040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0273781531654006e-05,
+      "loss": 3.8852,
+      "step": 1815552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0265395584143486e-05,
+      "loss": 3.8769,
+      "step": 1816064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0257009636632966e-05,
+      "loss": 3.8807,
+      "step": 1816576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0248623689122446e-05,
+      "loss": 3.8894,
+      "step": 1817088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0240237741611926e-05,
+      "loss": 3.8825,
+      "step": 1817600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0231868172905135e-05,
+      "loss": 3.8834,
+      "step": 1818112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0223482225394615e-05,
+      "loss": 3.8827,
+      "step": 1818624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0215096277884098e-05,
+      "loss": 3.8831,
+      "step": 1819136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0206710330373578e-05,
+      "loss": 3.8854,
+      "step": 1819648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0198324382863055e-05,
+      "loss": 3.8754,
+      "step": 1820160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0189954814156267e-05,
+      "loss": 3.8862,
+      "step": 1820672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0181568866645747e-05,
+      "loss": 3.874,
+      "step": 1821184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0173182919135224e-05,
+      "loss": 3.8976,
+      "step": 1821696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0164796971624707e-05,
+      "loss": 3.8847,
+      "step": 1822208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0156411024114187e-05,
+      "loss": 3.8882,
+      "step": 1822720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0148025076603667e-05,
+      "loss": 3.8759,
+      "step": 1823232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0139655507896876e-05,
+      "loss": 3.8892,
+      "step": 1823744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0131269560386356e-05,
+      "loss": 3.8889,
+      "step": 1824256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0122883612875836e-05,
+      "loss": 3.8878,
+      "step": 1824768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0114497665365316e-05,
+      "loss": 3.8803,
+      "step": 1825280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.01061117178548e-05,
+      "loss": 3.89,
+      "step": 1825792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.009772577034428e-05,
+      "loss": 3.885,
+      "step": 1826304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0089339822833756e-05,
+      "loss": 3.9002,
+      "step": 1826816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0080953875323236e-05,
+      "loss": 3.8771,
+      "step": 1827328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0072567927812716e-05,
+      "loss": 3.8934,
+      "step": 1827840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0064198359105928e-05,
+      "loss": 3.8871,
+      "step": 1828352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0055812411595408e-05,
+      "loss": 3.8869,
+      "step": 1828864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0047426464084888e-05,
+      "loss": 3.8852,
+      "step": 1829376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.00390568953781e-05,
+      "loss": 3.8892,
+      "step": 1829888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.003067094786758e-05,
+      "loss": 3.8893,
+      "step": 1830400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0022285000357057e-05,
+      "loss": 3.8705,
+      "step": 1830912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0013899052846537e-05,
+      "loss": 3.8921,
+      "step": 1831424
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.992622137069702,
+      "eval_runtime": 297.7714,
+      "eval_samples_per_second": 1281.49,
+      "eval_steps_per_second": 40.048,
+      "step": 1831680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0005545862943482e-05,
+      "loss": 3.8796,
+      "step": 1831936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9997159915432962e-05,
+      "loss": 3.8786,
+      "step": 1832448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9988773967922442e-05,
+      "loss": 3.8858,
+      "step": 1832960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9980388020411922e-05,
+      "loss": 3.8952,
+      "step": 1833472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9972002072901402e-05,
+      "loss": 3.8905,
+      "step": 1833984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9963616125390878e-05,
+      "loss": 3.8812,
+      "step": 1834496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.995523017788036e-05,
+      "loss": 3.8927,
+      "step": 1835008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.994684423036984e-05,
+      "loss": 3.8747,
+      "step": 1835520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.993845828285932e-05,
+      "loss": 3.8875,
+      "step": 1836032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.99300723353488e-05,
+      "loss": 3.8964,
+      "step": 1836544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.992168638783828e-05,
+      "loss": 3.885,
+      "step": 1837056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.991330044032776e-05,
+      "loss": 3.8873,
+      "step": 1837568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.990493087162097e-05,
+      "loss": 3.8902,
+      "step": 1838080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9896544924110454e-05,
+      "loss": 3.8745,
+      "step": 1838592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9888158976599934e-05,
+      "loss": 3.8752,
+      "step": 1839104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.987977302908941e-05,
+      "loss": 3.8743,
+      "step": 1839616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.987138708157889e-05,
+      "loss": 3.8826,
+      "step": 1840128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9863017512872103e-05,
+      "loss": 3.8834,
+      "step": 1840640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9854631565361583e-05,
+      "loss": 3.8807,
+      "step": 1841152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9846245617851063e-05,
+      "loss": 3.8911,
+      "step": 1841664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9837876049144275e-05,
+      "loss": 3.897,
+      "step": 1842176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9829490101633755e-05,
+      "loss": 3.893,
+      "step": 1842688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9821104154123235e-05,
+      "loss": 3.8861,
+      "step": 1843200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.981271820661271e-05,
+      "loss": 3.8858,
+      "step": 1843712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.980433225910219e-05,
+      "loss": 3.8857,
+      "step": 1844224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.979594631159167e-05,
+      "loss": 3.8864,
+      "step": 1844736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9787560364081155e-05,
+      "loss": 3.8865,
+      "step": 1845248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9779174416570635e-05,
+      "loss": 3.8755,
+      "step": 1845760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9770788469060115e-05,
+      "loss": 3.8822,
+      "step": 1846272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.976240252154959e-05,
+      "loss": 3.8758,
+      "step": 1846784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.975401657403907e-05,
+      "loss": 3.8791,
+      "step": 1847296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.974563062652855e-05,
+      "loss": 3.8904,
+      "step": 1847808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9737261057821764e-05,
+      "loss": 3.8904,
+      "step": 1848320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9728875110311244e-05,
+      "loss": 3.8843,
+      "step": 1848832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9720489162800724e-05,
+      "loss": 3.8803,
+      "step": 1849344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9712119594093936e-05,
+      "loss": 3.8914,
+      "step": 1849856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9703733646583416e-05,
+      "loss": 3.8829,
+      "step": 1850368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9695347699072893e-05,
+      "loss": 3.8767,
+      "step": 1850880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9686961751562376e-05,
+      "loss": 3.8809,
+      "step": 1851392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9678575804051856e-05,
+      "loss": 3.8715,
+      "step": 1851904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9670206235345065e-05,
+      "loss": 3.8823,
+      "step": 1852416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9661820287834545e-05,
+      "loss": 3.8903,
+      "step": 1852928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9653434340324025e-05,
+      "loss": 3.8792,
+      "step": 1853440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9645048392813505e-05,
+      "loss": 3.8815,
+      "step": 1853952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9636678824106717e-05,
+      "loss": 3.8948,
+      "step": 1854464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9628292876596197e-05,
+      "loss": 3.8863,
+      "step": 1854976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9619906929085677e-05,
+      "loss": 3.8884,
+      "step": 1855488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9611520981575157e-05,
+      "loss": 3.8805,
+      "step": 1856000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9603135034064637e-05,
+      "loss": 3.8564,
+      "step": 1856512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9594765465357846e-05,
+      "loss": 3.8981,
+      "step": 1857024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9586379517847326e-05,
+      "loss": 3.8853,
+      "step": 1857536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.957800994914054e-05,
+      "loss": 3.8867,
+      "step": 1858048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.956962400163002e-05,
+      "loss": 3.8836,
+      "step": 1858560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.95612380541195e-05,
+      "loss": 3.8728,
+      "step": 1859072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.955285210660898e-05,
+      "loss": 3.8686,
+      "step": 1859584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.954446615909846e-05,
+      "loss": 3.8775,
+      "step": 1860096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.953608021158794e-05,
+      "loss": 3.8725,
+      "step": 1860608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.952769426407742e-05,
+      "loss": 3.8737,
+      "step": 1861120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.95193083165669e-05,
+      "loss": 3.8947,
+      "step": 1861632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951092236905638e-05,
+      "loss": 3.8746,
+      "step": 1862144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.950253642154586e-05,
+      "loss": 3.8692,
+      "step": 1862656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.949415047403534e-05,
+      "loss": 3.8923,
+      "step": 1863168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.948576452652482e-05,
+      "loss": 3.8625,
+      "step": 1863680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.947739495781803e-05,
+      "loss": 3.8721,
+      "step": 1864192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.946900901030751e-05,
+      "loss": 3.8848,
+      "step": 1864704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.946063944160072e-05,
+      "loss": 3.8829,
+      "step": 1865216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.94522534940902e-05,
+      "loss": 3.8712,
+      "step": 1865728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9443883925383412e-05,
+      "loss": 3.876,
+      "step": 1866240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9435497977872892e-05,
+      "loss": 3.8681,
+      "step": 1866752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9427112030362372e-05,
+      "loss": 3.8843,
+      "step": 1867264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9418726082851852e-05,
+      "loss": 3.8822,
+      "step": 1867776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9410340135341332e-05,
+      "loss": 3.8776,
+      "step": 1868288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9401954187830812e-05,
+      "loss": 3.8811,
+      "step": 1868800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9393568240320292e-05,
+      "loss": 3.8847,
+      "step": 1869312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9385182292809772e-05,
+      "loss": 3.8939,
+      "step": 1869824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.937679634529925e-05,
+      "loss": 3.8693,
+      "step": 1870336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9368410397788732e-05,
+      "loss": 3.8846,
+      "step": 1870848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9360024450278212e-05,
+      "loss": 3.8873,
+      "step": 1871360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9351638502767692e-05,
+      "loss": 3.8641,
+      "step": 1871872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9343252555257172e-05,
+      "loss": 3.884,
+      "step": 1872384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9334899365354113e-05,
+      "loss": 3.8834,
+      "step": 1872896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9326529796647326e-05,
+      "loss": 3.8919,
+      "step": 1873408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9318143849136806e-05,
+      "loss": 3.8841,
+      "step": 1873920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9309757901626286e-05,
+      "loss": 3.8687,
+      "step": 1874432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9301371954115766e-05,
+      "loss": 3.861,
+      "step": 1874944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9292986006605246e-05,
+      "loss": 3.8745,
+      "step": 1875456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9284600059094722e-05,
+      "loss": 3.8857,
+      "step": 1875968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9276214111584202e-05,
+      "loss": 3.8777,
+      "step": 1876480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9267828164073686e-05,
+      "loss": 3.8738,
+      "step": 1876992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9259442216563166e-05,
+      "loss": 3.8758,
+      "step": 1877504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9251072647856375e-05,
+      "loss": 3.8851,
+      "step": 1878016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9242686700345855e-05,
+      "loss": 3.8743,
+      "step": 1878528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9234300752835335e-05,
+      "loss": 3.8735,
+      "step": 1879040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9225914805324815e-05,
+      "loss": 3.8816,
+      "step": 1879552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9217528857814294e-05,
+      "loss": 3.8895,
+      "step": 1880064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9209159289107507e-05,
+      "loss": 3.881,
+      "step": 1880576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9200773341596987e-05,
+      "loss": 3.8706,
+      "step": 1881088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9192387394086467e-05,
+      "loss": 3.8745,
+      "step": 1881600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9184001446575947e-05,
+      "loss": 3.8752,
+      "step": 1882112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9175631877869156e-05,
+      "loss": 3.8861,
+      "step": 1882624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.916724593035864e-05,
+      "loss": 3.8869,
+      "step": 1883136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.915885998284812e-05,
+      "loss": 3.8787,
+      "step": 1883648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.91504740353376e-05,
+      "loss": 3.876,
+      "step": 1884160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9142104466630808e-05,
+      "loss": 3.871,
+      "step": 1884672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9133718519120288e-05,
+      "loss": 3.8726,
+      "step": 1885184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9125332571609768e-05,
+      "loss": 3.8771,
+      "step": 1885696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9116946624099248e-05,
+      "loss": 3.8641,
+      "step": 1886208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.910856067658873e-05,
+      "loss": 3.8873,
+      "step": 1886720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9100174729078208e-05,
+      "loss": 3.8723,
+      "step": 1887232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9091788781567688e-05,
+      "loss": 3.8767,
+      "step": 1887744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.90834192128609e-05,
+      "loss": 3.8717,
+      "step": 1888256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9075033265350377e-05,
+      "loss": 3.869,
+      "step": 1888768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9066647317839857e-05,
+      "loss": 3.8877,
+      "step": 1889280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.905826137032934e-05,
+      "loss": 3.8702,
+      "step": 1889792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.904987542281882e-05,
+      "loss": 3.8958,
+      "step": 1890304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.90414894753083e-05,
+      "loss": 3.8861,
+      "step": 1890816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.903311990660151e-05,
+      "loss": 3.8803,
+      "step": 1891328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9024750337894722e-05,
+      "loss": 3.877,
+      "step": 1891840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9016364390384202e-05,
+      "loss": 3.8752,
+      "step": 1892352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9007978442873682e-05,
+      "loss": 3.8761,
+      "step": 1892864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8999592495363162e-05,
+      "loss": 3.8854,
+      "step": 1893376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.899120654785264e-05,
+      "loss": 3.8749,
+      "step": 1893888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.898282060034212e-05,
+      "loss": 3.8812,
+      "step": 1894400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.89744346528316e-05,
+      "loss": 3.8732,
+      "step": 1894912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.896604870532108e-05,
+      "loss": 3.8782,
+      "step": 1895424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8957662757810558e-05,
+      "loss": 3.8787,
+      "step": 1895936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.894927681030004e-05,
+      "loss": 3.8687,
+      "step": 1896448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8940907241593254e-05,
+      "loss": 3.8813,
+      "step": 1896960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8932521294082734e-05,
+      "loss": 3.8704,
+      "step": 1897472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.892413534657221e-05,
+      "loss": 3.8924,
+      "step": 1897984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.891574939906169e-05,
+      "loss": 3.8795,
+      "step": 1898496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.890736345155117e-05,
+      "loss": 3.8817,
+      "step": 1899008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.889897750404065e-05,
+      "loss": 3.869,
+      "step": 1899520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8890591556530134e-05,
+      "loss": 3.8896,
+      "step": 1900032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8882221987823343e-05,
+      "loss": 3.8773,
+      "step": 1900544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8873836040312823e-05,
+      "loss": 3.882,
+      "step": 1901056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8865466471606032e-05,
+      "loss": 3.8751,
+      "step": 1901568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8857080524095512e-05,
+      "loss": 3.8846,
+      "step": 1902080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8848694576584995e-05,
+      "loss": 3.8797,
+      "step": 1902592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8840308629074475e-05,
+      "loss": 3.8956,
+      "step": 1903104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8831922681563955e-05,
+      "loss": 3.8704,
+      "step": 1903616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8823553112857164e-05,
+      "loss": 3.8868,
+      "step": 1904128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8815167165346644e-05,
+      "loss": 3.883,
+      "step": 1904640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8806781217836124e-05,
+      "loss": 3.8842,
+      "step": 1905152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8798395270325604e-05,
+      "loss": 3.8744,
+      "step": 1905664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8790009322815087e-05,
+      "loss": 3.886,
+      "step": 1906176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8781623375304564e-05,
+      "loss": 3.8824,
+      "step": 1906688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8773253806597776e-05,
+      "loss": 3.8734,
+      "step": 1907200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8764867859087256e-05,
+      "loss": 3.8817,
+      "step": 1907712
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9909427165985107,
+      "eval_runtime": 291.1556,
+      "eval_samples_per_second": 1310.608,
+      "eval_steps_per_second": 40.957,
+      "step": 1908000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8756481911576736e-05,
+      "loss": 3.8743,
+      "step": 1908224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8748095964066216e-05,
+      "loss": 3.8704,
+      "step": 1908736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8739710016555696e-05,
+      "loss": 3.8792,
+      "step": 1909248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8731324069045176e-05,
+      "loss": 3.8906,
+      "step": 1909760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8722938121534656e-05,
+      "loss": 3.8863,
+      "step": 1910272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8714552174024136e-05,
+      "loss": 3.8738,
+      "step": 1910784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8706166226513616e-05,
+      "loss": 3.89,
+      "step": 1911296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8697780279003096e-05,
+      "loss": 3.8679,
+      "step": 1911808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8689394331492573e-05,
+      "loss": 3.8786,
+      "step": 1912320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8681008383982056e-05,
+      "loss": 3.8925,
+      "step": 1912832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8672622436471536e-05,
+      "loss": 3.8766,
+      "step": 1913344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8664236488961016e-05,
+      "loss": 3.8829,
+      "step": 1913856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8655866920254225e-05,
+      "loss": 3.8851,
+      "step": 1914368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8647480972743705e-05,
+      "loss": 3.8724,
+      "step": 1914880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8639095025233185e-05,
+      "loss": 3.8702,
+      "step": 1915392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8630709077722665e-05,
+      "loss": 3.8696,
+      "step": 1915904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8622323130212148e-05,
+      "loss": 3.875,
+      "step": 1916416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8613937182701628e-05,
+      "loss": 3.8771,
+      "step": 1916928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8605551235191105e-05,
+      "loss": 3.8741,
+      "step": 1917440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8597165287680585e-05,
+      "loss": 3.8877,
+      "step": 1917952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8588795718973797e-05,
+      "loss": 3.8912,
+      "step": 1918464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8580409771463277e-05,
+      "loss": 3.8901,
+      "step": 1918976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8572023823952757e-05,
+      "loss": 3.8827,
+      "step": 1919488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8563637876442237e-05,
+      "loss": 3.8765,
+      "step": 1920000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8555251928931717e-05,
+      "loss": 3.8839,
+      "step": 1920512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8546865981421197e-05,
+      "loss": 3.879,
+      "step": 1921024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8538496412714406e-05,
+      "loss": 3.8763,
+      "step": 1921536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8530110465203886e-05,
+      "loss": 3.8748,
+      "step": 1922048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8521724517693366e-05,
+      "loss": 3.8759,
+      "step": 1922560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.851333857018285e-05,
+      "loss": 3.8728,
+      "step": 1923072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.850495262267233e-05,
+      "loss": 3.8735,
+      "step": 1923584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.849656667516181e-05,
+      "loss": 3.8821,
+      "step": 1924096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8488180727651286e-05,
+      "loss": 3.8886,
+      "step": 1924608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8479794780140766e-05,
+      "loss": 3.8796,
+      "step": 1925120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8471425211433978e-05,
+      "loss": 3.8774,
+      "step": 1925632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.846305564272719e-05,
+      "loss": 3.8781,
+      "step": 1926144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.845466969521667e-05,
+      "loss": 3.8837,
+      "step": 1926656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.844628374770615e-05,
+      "loss": 3.87,
+      "step": 1927168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.843789780019563e-05,
+      "loss": 3.8768,
+      "step": 1927680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8429511852685107e-05,
+      "loss": 3.8655,
+      "step": 1928192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8421125905174587e-05,
+      "loss": 3.8797,
+      "step": 1928704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8412756336467803e-05,
+      "loss": 3.881,
+      "step": 1929216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8404370388957283e-05,
+      "loss": 3.8744,
+      "step": 1929728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.839598444144676e-05,
+      "loss": 3.8793,
+      "step": 1930240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.838759849393624e-05,
+      "loss": 3.8871,
+      "step": 1930752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.837921254642572e-05,
+      "loss": 3.8824,
+      "step": 1931264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.83708265989152e-05,
+      "loss": 3.8834,
+      "step": 1931776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.836244065140468e-05,
+      "loss": 3.8771,
+      "step": 1932288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8354071082697892e-05,
+      "loss": 3.8479,
+      "step": 1932800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8345701513991104e-05,
+      "loss": 3.8876,
+      "step": 1933312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.833731556648058e-05,
+      "loss": 3.8799,
+      "step": 1933824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.832892961897006e-05,
+      "loss": 3.883,
+      "step": 1934336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.832054367145954e-05,
+      "loss": 3.8792,
+      "step": 1934848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.831215772394902e-05,
+      "loss": 3.8689,
+      "step": 1935360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8303771776438504e-05,
+      "loss": 3.8648,
+      "step": 1935872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8295385828927984e-05,
+      "loss": 3.8704,
+      "step": 1936384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8286999881417464e-05,
+      "loss": 3.8717,
+      "step": 1936896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8278630312710673e-05,
+      "loss": 3.867,
+      "step": 1937408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8270244365200153e-05,
+      "loss": 3.8895,
+      "step": 1937920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8261858417689633e-05,
+      "loss": 3.8731,
+      "step": 1938432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8253472470179113e-05,
+      "loss": 3.8558,
+      "step": 1938944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8245086522668593e-05,
+      "loss": 3.8899,
+      "step": 1939456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8236700575158073e-05,
+      "loss": 3.862,
+      "step": 1939968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8228331006451285e-05,
+      "loss": 3.8638,
+      "step": 1940480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8219945058940762e-05,
+      "loss": 3.8807,
+      "step": 1940992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8211559111430242e-05,
+      "loss": 3.8748,
+      "step": 1941504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8203173163919722e-05,
+      "loss": 3.8689,
+      "step": 1942016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8194787216409205e-05,
+      "loss": 3.8687,
+      "step": 1942528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8186401268898685e-05,
+      "loss": 3.863,
+      "step": 1943040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8178015321388165e-05,
+      "loss": 3.8749,
+      "step": 1943552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8169629373877645e-05,
+      "loss": 3.8743,
+      "step": 1944064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8161259805170854e-05,
+      "loss": 3.8761,
+      "step": 1944576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8152873857660334e-05,
+      "loss": 3.875,
+      "step": 1945088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8144487910149814e-05,
+      "loss": 3.8787,
+      "step": 1945600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8136101962639294e-05,
+      "loss": 3.889,
+      "step": 1946112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8127716015128774e-05,
+      "loss": 3.8617,
+      "step": 1946624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8119330067618254e-05,
+      "loss": 3.8784,
+      "step": 1947136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8110944120107734e-05,
+      "loss": 3.8851,
+      "step": 1947648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8102558172597214e-05,
+      "loss": 3.8615,
+      "step": 1948160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8094188603890426e-05,
+      "loss": 3.8729,
+      "step": 1948672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8085802656379906e-05,
+      "loss": 3.881,
+      "step": 1949184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8077416708869386e-05,
+      "loss": 3.8845,
+      "step": 1949696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8069030761358866e-05,
+      "loss": 3.8833,
+      "step": 1950208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8060661192652075e-05,
+      "loss": 3.8618,
+      "step": 1950720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8052291623945288e-05,
+      "loss": 3.8551,
+      "step": 1951232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8043905676434768e-05,
+      "loss": 3.8758,
+      "step": 1951744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8035519728924248e-05,
+      "loss": 3.8776,
+      "step": 1952256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8027133781413728e-05,
+      "loss": 3.8692,
+      "step": 1952768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.801876421270694e-05,
+      "loss": 3.8724,
+      "step": 1953280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8010378265196417e-05,
+      "loss": 3.8686,
+      "step": 1953792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8001992317685897e-05,
+      "loss": 3.88,
+      "step": 1954304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.799360637017538e-05,
+      "loss": 3.873,
+      "step": 1954816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.798522042266486e-05,
+      "loss": 3.8687,
+      "step": 1955328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.797683447515434e-05,
+      "loss": 3.872,
+      "step": 1955840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.796844852764382e-05,
+      "loss": 3.8836,
+      "step": 1956352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.796007895893703e-05,
+      "loss": 3.878,
+      "step": 1956864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.795169301142651e-05,
+      "loss": 3.8658,
+      "step": 1957376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.794330706391599e-05,
+      "loss": 3.8674,
+      "step": 1957888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.793492111640547e-05,
+      "loss": 3.8698,
+      "step": 1958400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.792653516889495e-05,
+      "loss": 3.8808,
+      "step": 1958912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.791814922138443e-05,
+      "loss": 3.8775,
+      "step": 1959424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.790976327387391e-05,
+      "loss": 3.8726,
+      "step": 1959936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.790137732636339e-05,
+      "loss": 3.8719,
+      "step": 1960448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7893007757656598e-05,
+      "loss": 3.8644,
+      "step": 1960960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7884638188949813e-05,
+      "loss": 3.873,
+      "step": 1961472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7876252241439293e-05,
+      "loss": 3.8671,
+      "step": 1961984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7867882672732503e-05,
+      "loss": 3.8613,
+      "step": 1962496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7859496725221982e-05,
+      "loss": 3.8819,
+      "step": 1963008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7851110777711462e-05,
+      "loss": 3.8665,
+      "step": 1963520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7842724830200942e-05,
+      "loss": 3.8704,
+      "step": 1964032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7834338882690422e-05,
+      "loss": 3.8665,
+      "step": 1964544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7825952935179902e-05,
+      "loss": 3.8636,
+      "step": 1965056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7817566987669382e-05,
+      "loss": 3.8796,
+      "step": 1965568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7809181040158862e-05,
+      "loss": 3.8677,
+      "step": 1966080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7800795092648342e-05,
+      "loss": 3.8882,
+      "step": 1966592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7792409145137822e-05,
+      "loss": 3.8817,
+      "step": 1967104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7784023197627302e-05,
+      "loss": 3.874,
+      "step": 1967616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7775653628920515e-05,
+      "loss": 3.8733,
+      "step": 1968128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7767267681409995e-05,
+      "loss": 3.8715,
+      "step": 1968640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7758881733899474e-05,
+      "loss": 3.866,
+      "step": 1969152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7750495786388954e-05,
+      "loss": 3.8811,
+      "step": 1969664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7742126217682164e-05,
+      "loss": 3.8754,
+      "step": 1970176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7733740270171644e-05,
+      "loss": 3.872,
+      "step": 1970688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7725354322661127e-05,
+      "loss": 3.8669,
+      "step": 1971200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7716968375150603e-05,
+      "loss": 3.8732,
+      "step": 1971712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7708582427640083e-05,
+      "loss": 3.8707,
+      "step": 1972224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7700196480129563e-05,
+      "loss": 3.8656,
+      "step": 1972736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7691826911422776e-05,
+      "loss": 3.8766,
+      "step": 1973248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7683440963912252e-05,
+      "loss": 3.8557,
+      "step": 1973760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7675055016401736e-05,
+      "loss": 3.8925,
+      "step": 1974272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7666669068891216e-05,
+      "loss": 3.8715,
+      "step": 1974784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7658283121380696e-05,
+      "loss": 3.8766,
+      "step": 1975296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7649897173870176e-05,
+      "loss": 3.8633,
+      "step": 1975808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7641527605163385e-05,
+      "loss": 3.8836,
+      "step": 1976320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7633141657652865e-05,
+      "loss": 3.8741,
+      "step": 1976832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7624755710142345e-05,
+      "loss": 3.8764,
+      "step": 1977344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7616369762631828e-05,
+      "loss": 3.871,
+      "step": 1977856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7607983815121308e-05,
+      "loss": 3.8749,
+      "step": 1978368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7599597867610784e-05,
+      "loss": 3.8739,
+      "step": 1978880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7591211920100264e-05,
+      "loss": 3.892,
+      "step": 1979392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7582825972589744e-05,
+      "loss": 3.8654,
+      "step": 1979904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7574440025079224e-05,
+      "loss": 3.882,
+      "step": 1980416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7566070456372437e-05,
+      "loss": 3.8791,
+      "step": 1980928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.755770088766565e-05,
+      "loss": 3.8793,
+      "step": 1981440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.754933131895886e-05,
+      "loss": 3.8682,
+      "step": 1981952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.754094537144834e-05,
+      "loss": 3.8793,
+      "step": 1982464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7532559423937818e-05,
+      "loss": 3.8753,
+      "step": 1982976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7524173476427298e-05,
+      "loss": 3.8687,
+      "step": 1983488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.751578752891678e-05,
+      "loss": 3.8764,
+      "step": 1984000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9886820316314697,
+      "eval_runtime": 292.2561,
+      "eval_samples_per_second": 1305.673,
+      "eval_steps_per_second": 40.803,
+      "step": 1984320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7507401581406258e-05,
+      "loss": 3.8663,
+      "step": 1984512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7499015633895738e-05,
+      "loss": 3.8618,
+      "step": 1985024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7490629686385218e-05,
+      "loss": 3.8753,
+      "step": 1985536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7482243738874698e-05,
+      "loss": 3.8837,
+      "step": 1986048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7473857791364178e-05,
+      "loss": 3.8832,
+      "step": 1986560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.746548822265739e-05,
+      "loss": 3.8714,
+      "step": 1987072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.745710227514687e-05,
+      "loss": 3.8786,
+      "step": 1987584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.744871632763635e-05,
+      "loss": 3.8677,
+      "step": 1988096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.744034675892956e-05,
+      "loss": 3.8718,
+      "step": 1988608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.743196081141904e-05,
+      "loss": 3.8862,
+      "step": 1989120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.742357486390852e-05,
+      "loss": 3.8736,
+      "step": 1989632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7415188916398e-05,
+      "loss": 3.8798,
+      "step": 1990144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7406819347691212e-05,
+      "loss": 3.8826,
+      "step": 1990656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7398433400180692e-05,
+      "loss": 3.8643,
+      "step": 1991168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7390047452670172e-05,
+      "loss": 3.865,
+      "step": 1991680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7381661505159652e-05,
+      "loss": 3.861,
+      "step": 1992192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.737327555764913e-05,
+      "loss": 3.8726,
+      "step": 1992704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.736488961013861e-05,
+      "loss": 3.8759,
+      "step": 1993216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.735650366262809e-05,
+      "loss": 3.8655,
+      "step": 1993728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.734811771511757e-05,
+      "loss": 3.8798,
+      "step": 1994240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.733973176760705e-05,
+      "loss": 3.886,
+      "step": 1994752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.733134582009653e-05,
+      "loss": 3.8829,
+      "step": 1995264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.732295987258601e-05,
+      "loss": 3.8799,
+      "step": 1995776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.731457392507549e-05,
+      "loss": 3.8717,
+      "step": 1996288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.73062043563687e-05,
+      "loss": 3.8787,
+      "step": 1996800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7297818408858184e-05,
+      "loss": 3.8756,
+      "step": 1997312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7289432461347664e-05,
+      "loss": 3.869,
+      "step": 1997824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7281046513837144e-05,
+      "loss": 3.8664,
+      "step": 1998336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.727266056632662e-05,
+      "loss": 3.8742,
+      "step": 1998848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.7264307376423562e-05,
+      "loss": 3.8678,
+      "step": 1999360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7255921428913045e-05,
+      "loss": 3.8686,
+      "step": 1999872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7247535481402525e-05,
+      "loss": 3.8744,
+      "step": 2000384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7239149533892005e-05,
+      "loss": 3.8784,
+      "step": 2000896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7230763586381485e-05,
+      "loss": 3.8799,
+      "step": 2001408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7222394017674694e-05,
+      "loss": 3.8647,
+      "step": 2001920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7214024448967907e-05,
+      "loss": 3.8765,
+      "step": 2002432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7205638501457387e-05,
+      "loss": 3.8778,
+      "step": 2002944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7197252553946867e-05,
+      "loss": 3.8596,
+      "step": 2003456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7188866606436347e-05,
+      "loss": 3.8718,
+      "step": 2003968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.718049703772956e-05,
+      "loss": 3.8606,
+      "step": 2004480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7172111090219036e-05,
+      "loss": 3.873,
+      "step": 2004992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7163725142708516e-05,
+      "loss": 3.8727,
+      "step": 2005504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7155339195198e-05,
+      "loss": 3.869,
+      "step": 2006016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.714695324768748e-05,
+      "loss": 3.8729,
+      "step": 2006528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.713856730017696e-05,
+      "loss": 3.8823,
+      "step": 2007040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.713018135266644e-05,
+      "loss": 3.8772,
+      "step": 2007552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7121795405155915e-05,
+      "loss": 3.8802,
+      "step": 2008064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7113409457645395e-05,
+      "loss": 3.8695,
+      "step": 2008576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7105023510134875e-05,
+      "loss": 3.8461,
+      "step": 2009088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7096637562624355e-05,
+      "loss": 3.8831,
+      "step": 2009600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.70882843727213e-05,
+      "loss": 3.8741,
+      "step": 2010112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.707989842521078e-05,
+      "loss": 3.8786,
+      "step": 2010624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.707151247770026e-05,
+      "loss": 3.8717,
+      "step": 2011136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7063126530189737e-05,
+      "loss": 3.8643,
+      "step": 2011648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.705474058267922e-05,
+      "loss": 3.8565,
+      "step": 2012160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.70463546351687e-05,
+      "loss": 3.8635,
+      "step": 2012672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.703796868765818e-05,
+      "loss": 3.8685,
+      "step": 2013184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.702958274014766e-05,
+      "loss": 3.8567,
+      "step": 2013696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.702119679263714e-05,
+      "loss": 3.8844,
+      "step": 2014208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.701281084512662e-05,
+      "loss": 3.8685,
+      "step": 2014720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7004424897616096e-05,
+      "loss": 3.8537,
+      "step": 2015232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.699605532890931e-05,
+      "loss": 3.8804,
+      "step": 2015744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6987669381398792e-05,
+      "loss": 3.8571,
+      "step": 2016256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6979283433888272e-05,
+      "loss": 3.8592,
+      "step": 2016768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.697089748637775e-05,
+      "loss": 3.8722,
+      "step": 2017280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.696251153886723e-05,
+      "loss": 3.8679,
+      "step": 2017792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.695412559135671e-05,
+      "loss": 3.8641,
+      "step": 2018304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.694573964384619e-05,
+      "loss": 3.8631,
+      "step": 2018816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.693735369633567e-05,
+      "loss": 3.8615,
+      "step": 2019328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.692898412762888e-05,
+      "loss": 3.869,
+      "step": 2019840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.692059818011836e-05,
+      "loss": 3.8764,
+      "step": 2020352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.691221223260784e-05,
+      "loss": 3.8685,
+      "step": 2020864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.690384266390105e-05,
+      "loss": 3.8704,
+      "step": 2021376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.689545671639053e-05,
+      "loss": 3.8682,
+      "step": 2021888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6887070768880013e-05,
+      "loss": 3.8857,
+      "step": 2022400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6878684821369493e-05,
+      "loss": 3.8606,
+      "step": 2022912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6870298873858973e-05,
+      "loss": 3.875,
+      "step": 2023424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6861912926348453e-05,
+      "loss": 3.8736,
+      "step": 2023936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.685352697883793e-05,
+      "loss": 3.8612,
+      "step": 2024448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.684514103132741e-05,
+      "loss": 3.8677,
+      "step": 2024960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6836771462620622e-05,
+      "loss": 3.8749,
+      "step": 2025472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6828385515110102e-05,
+      "loss": 3.8783,
+      "step": 2025984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6819999567599582e-05,
+      "loss": 3.8796,
+      "step": 2026496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6811613620089062e-05,
+      "loss": 3.8499,
+      "step": 2027008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6803244051382275e-05,
+      "loss": 3.8548,
+      "step": 2027520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.679485810387175e-05,
+      "loss": 3.8671,
+      "step": 2028032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.678647215636123e-05,
+      "loss": 3.8728,
+      "step": 2028544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6778102587654447e-05,
+      "loss": 3.8635,
+      "step": 2029056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6769716640143927e-05,
+      "loss": 3.8683,
+      "step": 2029568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6761347071437136e-05,
+      "loss": 3.865,
+      "step": 2030080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6752961123926616e-05,
+      "loss": 3.8745,
+      "step": 2030592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6744575176416096e-05,
+      "loss": 3.8666,
+      "step": 2031104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.673620560770931e-05,
+      "loss": 3.8621,
+      "step": 2031616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.672781966019879e-05,
+      "loss": 3.8697,
+      "step": 2032128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.671943371268827e-05,
+      "loss": 3.8781,
+      "step": 2032640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6711047765177748e-05,
+      "loss": 3.8742,
+      "step": 2033152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6702661817667225e-05,
+      "loss": 3.8571,
+      "step": 2033664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6694275870156705e-05,
+      "loss": 3.8617,
+      "step": 2034176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6685889922646185e-05,
+      "loss": 3.8655,
+      "step": 2034688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6677503975135668e-05,
+      "loss": 3.8758,
+      "step": 2035200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6669134406428877e-05,
+      "loss": 3.8744,
+      "step": 2035712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6660748458918357e-05,
+      "loss": 3.866,
+      "step": 2036224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6652362511407837e-05,
+      "loss": 3.8657,
+      "step": 2036736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6643976563897317e-05,
+      "loss": 3.8587,
+      "step": 2037248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6635590616386797e-05,
+      "loss": 3.8654,
+      "step": 2037760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6627204668876277e-05,
+      "loss": 3.8614,
+      "step": 2038272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6618818721365757e-05,
+      "loss": 3.8589,
+      "step": 2038784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6610432773855237e-05,
+      "loss": 3.8737,
+      "step": 2039296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6602046826344717e-05,
+      "loss": 3.8626,
+      "step": 2039808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6593660878834197e-05,
+      "loss": 3.8637,
+      "step": 2040320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6585274931323677e-05,
+      "loss": 3.8607,
+      "step": 2040832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6576888983813157e-05,
+      "loss": 3.8573,
+      "step": 2041344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6568503036302637e-05,
+      "loss": 3.8745,
+      "step": 2041856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.656013346759585e-05,
+      "loss": 3.868,
+      "step": 2042368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.655174752008533e-05,
+      "loss": 3.8776,
+      "step": 2042880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6543377951378538e-05,
+      "loss": 3.8749,
+      "step": 2043392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6534992003868018e-05,
+      "loss": 3.8727,
+      "step": 2043904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6526606056357498e-05,
+      "loss": 3.8682,
+      "step": 2044416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6518220108846978e-05,
+      "loss": 3.867,
+      "step": 2044928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.650985054014019e-05,
+      "loss": 3.8644,
+      "step": 2045440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.650146459262967e-05,
+      "loss": 3.8747,
+      "step": 2045952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.649307864511915e-05,
+      "loss": 3.8678,
+      "step": 2046464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.648469269760863e-05,
+      "loss": 3.8734,
+      "step": 2046976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.647630675009811e-05,
+      "loss": 3.8587,
+      "step": 2047488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6467937181391323e-05,
+      "loss": 3.8714,
+      "step": 2048000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6459551233880803e-05,
+      "loss": 3.8639,
+      "step": 2048512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6451165286370283e-05,
+      "loss": 3.8602,
+      "step": 2049024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.644277933885976e-05,
+      "loss": 3.8706,
+      "step": 2049536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.643439339134924e-05,
+      "loss": 3.8518,
+      "step": 2050048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.642600744383872e-05,
+      "loss": 3.8884,
+      "step": 2050560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.64176214963282e-05,
+      "loss": 3.8671,
+      "step": 2051072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.640923554881768e-05,
+      "loss": 3.8745,
+      "step": 2051584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.640086598011089e-05,
+      "loss": 3.8592,
+      "step": 2052096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.639248003260037e-05,
+      "loss": 3.8761,
+      "step": 2052608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.638409408508985e-05,
+      "loss": 3.8685,
+      "step": 2053120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.637570813757933e-05,
+      "loss": 3.8735,
+      "step": 2053632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.636732219006881e-05,
+      "loss": 3.8666,
+      "step": 2054144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6358952621362024e-05,
+      "loss": 3.8665,
+      "step": 2054656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6350566673851504e-05,
+      "loss": 3.8717,
+      "step": 2055168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6342180726340984e-05,
+      "loss": 3.8906,
+      "step": 2055680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6333794778830464e-05,
+      "loss": 3.8556,
+      "step": 2056192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.632540883131994e-05,
+      "loss": 3.8808,
+      "step": 2056704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6317039262613153e-05,
+      "loss": 3.8761,
+      "step": 2057216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6308653315102633e-05,
+      "loss": 3.8684,
+      "step": 2057728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6300283746395845e-05,
+      "loss": 3.863,
+      "step": 2058240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6291897798885325e-05,
+      "loss": 3.873,
+      "step": 2058752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6283528230178534e-05,
+      "loss": 3.8691,
+      "step": 2059264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6275142282668014e-05,
+      "loss": 3.8678,
+      "step": 2059776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6266756335157494e-05,
+      "loss": 3.867,
+      "step": 2060288
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.986829996109009,
+      "eval_runtime": 289.8649,
+      "eval_samples_per_second": 1316.444,
+      "eval_steps_per_second": 41.14,
+      "step": 2060640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6258370387646978e-05,
+      "loss": 3.8608,
+      "step": 2060800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6249984440136458e-05,
+      "loss": 3.8597,
+      "step": 2061312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6241598492625938e-05,
+      "loss": 3.8678,
+      "step": 2061824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6233212545115414e-05,
+      "loss": 3.8776,
+      "step": 2062336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6224826597604894e-05,
+      "loss": 3.8799,
+      "step": 2062848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6216440650094374e-05,
+      "loss": 3.8664,
+      "step": 2063360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6208054702583854e-05,
+      "loss": 3.8762,
+      "step": 2063872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6199668755073334e-05,
+      "loss": 3.8651,
+      "step": 2064384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6191282807562817e-05,
+      "loss": 3.8636,
+      "step": 2064896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6182896860052297e-05,
+      "loss": 3.877,
+      "step": 2065408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6174510912541774e-05,
+      "loss": 3.8758,
+      "step": 2065920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6166124965031254e-05,
+      "loss": 3.876,
+      "step": 2066432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6157755396324466e-05,
+      "loss": 3.8735,
+      "step": 2066944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6149369448813946e-05,
+      "loss": 3.8618,
+      "step": 2067456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6140983501303426e-05,
+      "loss": 3.8636,
+      "step": 2067968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6132597553792906e-05,
+      "loss": 3.8499,
+      "step": 2068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6124211606282386e-05,
+      "loss": 3.8706,
+      "step": 2068992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6115825658771866e-05,
+      "loss": 3.8705,
+      "step": 2069504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6107439711261346e-05,
+      "loss": 3.8604,
+      "step": 2070016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6099053763750826e-05,
+      "loss": 3.8726,
+      "step": 2070528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6090667816240302e-05,
+      "loss": 3.8816,
+      "step": 2071040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6082281868729786e-05,
+      "loss": 3.8751,
+      "step": 2071552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6073895921219266e-05,
+      "loss": 3.875,
+      "step": 2072064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6065509973708746e-05,
+      "loss": 3.8671,
+      "step": 2072576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6057124026198226e-05,
+      "loss": 3.8716,
+      "step": 2073088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6048754457491435e-05,
+      "loss": 3.8711,
+      "step": 2073600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6040368509980915e-05,
+      "loss": 3.8632,
+      "step": 2074112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.603199894127413e-05,
+      "loss": 3.8596,
+      "step": 2074624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6023612993763607e-05,
+      "loss": 3.8705,
+      "step": 2075136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6015227046253087e-05,
+      "loss": 3.8624,
+      "step": 2075648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6006841098742567e-05,
+      "loss": 3.8692,
+      "step": 2076160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5998455151232047e-05,
+      "loss": 3.8702,
+      "step": 2076672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5990085582525256e-05,
+      "loss": 3.8702,
+      "step": 2077184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.598169963501474e-05,
+      "loss": 3.8785,
+      "step": 2077696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5973330066307952e-05,
+      "loss": 3.8609,
+      "step": 2078208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.596496049760116e-05,
+      "loss": 3.8729,
+      "step": 2078720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.595657455009064e-05,
+      "loss": 3.8745,
+      "step": 2079232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.594818860258012e-05,
+      "loss": 3.8527,
+      "step": 2079744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.59398026550696e-05,
+      "loss": 3.8696,
+      "step": 2080256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.593141670755908e-05,
+      "loss": 3.8576,
+      "step": 2080768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.592303076004856e-05,
+      "loss": 3.8634,
+      "step": 2081280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.591464481253804e-05,
+      "loss": 3.8685,
+      "step": 2081792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.590625886502752e-05,
+      "loss": 3.8657,
+      "step": 2082304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5897872917517e-05,
+      "loss": 3.8701,
+      "step": 2082816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.588948697000648e-05,
+      "loss": 3.8757,
+      "step": 2083328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5881101022495957e-05,
+      "loss": 3.8735,
+      "step": 2083840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.587271507498544e-05,
+      "loss": 3.8734,
+      "step": 2084352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.586432912747492e-05,
+      "loss": 3.8674,
+      "step": 2084864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5855959558768133e-05,
+      "loss": 3.8428,
+      "step": 2085376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.584757361125761e-05,
+      "loss": 3.8779,
+      "step": 2085888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.583918766374709e-05,
+      "loss": 3.8695,
+      "step": 2086400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.583080171623657e-05,
+      "loss": 3.8768,
+      "step": 2086912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.582241576872605e-05,
+      "loss": 3.8675,
+      "step": 2087424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5814029821215533e-05,
+      "loss": 3.8559,
+      "step": 2087936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5805643873705013e-05,
+      "loss": 3.8603,
+      "step": 2088448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.579725792619449e-05,
+      "loss": 3.8534,
+      "step": 2088960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.578887197868397e-05,
+      "loss": 3.8656,
+      "step": 2089472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5780502409977182e-05,
+      "loss": 3.8497,
+      "step": 2089984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5772132841270394e-05,
+      "loss": 3.8813,
+      "step": 2090496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5763746893759874e-05,
+      "loss": 3.8666,
+      "step": 2091008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5755360946249354e-05,
+      "loss": 3.847,
+      "step": 2091520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5746974998738834e-05,
+      "loss": 3.8745,
+      "step": 2092032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5738589051228314e-05,
+      "loss": 3.8556,
+      "step": 2092544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.573020310371779e-05,
+      "loss": 3.8558,
+      "step": 2093056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5721833535011003e-05,
+      "loss": 3.8652,
+      "step": 2093568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5713447587500486e-05,
+      "loss": 3.8638,
+      "step": 2094080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5705061639989963e-05,
+      "loss": 3.8586,
+      "step": 2094592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5696675692479443e-05,
+      "loss": 3.8623,
+      "step": 2095104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5688289744968923e-05,
+      "loss": 3.8524,
+      "step": 2095616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5679903797458403e-05,
+      "loss": 3.8637,
+      "step": 2096128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5671517849947883e-05,
+      "loss": 3.8706,
+      "step": 2096640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5663148281241095e-05,
+      "loss": 3.8617,
+      "step": 2097152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5654762333730575e-05,
+      "loss": 3.869,
+      "step": 2097664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5646376386220055e-05,
+      "loss": 3.8596,
+      "step": 2098176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5638006817513264e-05,
+      "loss": 3.8852,
+      "step": 2098688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5629620870002744e-05,
+      "loss": 3.8526,
+      "step": 2099200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5621234922492224e-05,
+      "loss": 3.8721,
+      "step": 2099712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5612848974981704e-05,
+      "loss": 3.8597,
+      "step": 2100224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5604463027471188e-05,
+      "loss": 3.8633,
+      "step": 2100736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5596077079960668e-05,
+      "loss": 3.8599,
+      "step": 2101248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5587691132450144e-05,
+      "loss": 3.8686,
+      "step": 2101760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5579305184939624e-05,
+      "loss": 3.8734,
+      "step": 2102272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5570919237429104e-05,
+      "loss": 3.8742,
+      "step": 2102784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5562533289918584e-05,
+      "loss": 3.8482,
+      "step": 2103296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5554147342408064e-05,
+      "loss": 3.8507,
+      "step": 2103808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5545761394897547e-05,
+      "loss": 3.8617,
+      "step": 2104320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5537391826190756e-05,
+      "loss": 3.8709,
+      "step": 2104832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5529005878680236e-05,
+      "loss": 3.859,
+      "step": 2105344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5520636309973445e-05,
+      "loss": 3.8632,
+      "step": 2105856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5512250362462925e-05,
+      "loss": 3.862,
+      "step": 2106368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.550386441495241e-05,
+      "loss": 3.8635,
+      "step": 2106880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.549547846744189e-05,
+      "loss": 3.8649,
+      "step": 2107392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.548709251993137e-05,
+      "loss": 3.8602,
+      "step": 2107904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.547870657242085e-05,
+      "loss": 3.8659,
+      "step": 2108416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5470320624910325e-05,
+      "loss": 3.8721,
+      "step": 2108928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5461934677399805e-05,
+      "loss": 3.8746,
+      "step": 2109440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5453565108693018e-05,
+      "loss": 3.8527,
+      "step": 2109952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5445179161182498e-05,
+      "loss": 3.8549,
+      "step": 2110464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5436793213671977e-05,
+      "loss": 3.8627,
+      "step": 2110976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5428407266161457e-05,
+      "loss": 3.868,
+      "step": 2111488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5420021318650937e-05,
+      "loss": 3.872,
+      "step": 2112000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.541165174994415e-05,
+      "loss": 3.8619,
+      "step": 2112512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5403265802433626e-05,
+      "loss": 3.8634,
+      "step": 2113024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.539487985492311e-05,
+      "loss": 3.8555,
+      "step": 2113536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.538649390741259e-05,
+      "loss": 3.8595,
+      "step": 2114048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.537810795990207e-05,
+      "loss": 3.8599,
+      "step": 2114560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.536973839119528e-05,
+      "loss": 3.8543,
+      "step": 2115072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.536135244368476e-05,
+      "loss": 3.8652,
+      "step": 2115584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.535296649617424e-05,
+      "loss": 3.8587,
+      "step": 2116096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.534459692746745e-05,
+      "loss": 3.8579,
+      "step": 2116608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.533621097995693e-05,
+      "loss": 3.8588,
+      "step": 2117120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.532782503244641e-05,
+      "loss": 3.8512,
+      "step": 2117632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.531945546373962e-05,
+      "loss": 3.8699,
+      "step": 2118144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.53110695162291e-05,
+      "loss": 3.8639,
+      "step": 2118656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.530268356871858e-05,
+      "loss": 3.8709,
+      "step": 2119168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5294297621208063e-05,
+      "loss": 3.8646,
+      "step": 2119680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5285911673697543e-05,
+      "loss": 3.8722,
+      "step": 2120192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5277525726187023e-05,
+      "loss": 3.8659,
+      "step": 2120704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5269156157480232e-05,
+      "loss": 3.8548,
+      "step": 2121216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5260770209969712e-05,
+      "loss": 3.8656,
+      "step": 2121728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5252384262459194e-05,
+      "loss": 3.865,
+      "step": 2122240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5243998314948674e-05,
+      "loss": 3.8612,
+      "step": 2122752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5235612367438154e-05,
+      "loss": 3.8702,
+      "step": 2123264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5227226419927632e-05,
+      "loss": 3.8584,
+      "step": 2123776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5218840472417112e-05,
+      "loss": 3.8631,
+      "step": 2124288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5210454524906592e-05,
+      "loss": 3.8565,
+      "step": 2124800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5202068577396072e-05,
+      "loss": 3.8589,
+      "step": 2125312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5193699008689283e-05,
+      "loss": 3.8639,
+      "step": 2125824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5185329439982495e-05,
+      "loss": 3.8525,
+      "step": 2126336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5176943492471975e-05,
+      "loss": 3.8805,
+      "step": 2126848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5168557544961454e-05,
+      "loss": 3.8629,
+      "step": 2127360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5160171597450934e-05,
+      "loss": 3.8681,
+      "step": 2127872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5151785649940414e-05,
+      "loss": 3.8538,
+      "step": 2128384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5143399702429895e-05,
+      "loss": 3.8691,
+      "step": 2128896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5135013754919375e-05,
+      "loss": 3.8656,
+      "step": 2129408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5126627807408855e-05,
+      "loss": 3.868,
+      "step": 2129920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5118241859898335e-05,
+      "loss": 3.8614,
+      "step": 2130432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5109855912387813e-05,
+      "loss": 3.8608,
+      "step": 2130944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5101469964877293e-05,
+      "loss": 3.8651,
+      "step": 2131456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5093084017366773e-05,
+      "loss": 3.8844,
+      "step": 2131968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5084714448659984e-05,
+      "loss": 3.8553,
+      "step": 2132480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5076344879953196e-05,
+      "loss": 3.8735,
+      "step": 2132992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5067958932442678e-05,
+      "loss": 3.8732,
+      "step": 2133504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5059572984932158e-05,
+      "loss": 3.8622,
+      "step": 2134016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5051187037421635e-05,
+      "loss": 3.8672,
+      "step": 2134528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5042801089911116e-05,
+      "loss": 3.8651,
+      "step": 2135040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5034431521204329e-05,
+      "loss": 3.8654,
+      "step": 2135552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5026045573693809e-05,
+      "loss": 3.864,
+      "step": 2136064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5017659626183287e-05,
+      "loss": 3.8619,
+      "step": 2136576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9857234954833984,
+      "eval_runtime": 287.3327,
+      "eval_samples_per_second": 1328.046,
+      "eval_steps_per_second": 41.502,
+      "step": 2136960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5009273678672767e-05,
+      "loss": 3.8614,
+      "step": 2137088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5000887731162247e-05,
+      "loss": 3.849,
+      "step": 2137600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4992501783651727e-05,
+      "loss": 3.8597,
+      "step": 2138112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4984115836141207e-05,
+      "loss": 3.8761,
+      "step": 2138624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4975729888630688e-05,
+      "loss": 3.8707,
+      "step": 2139136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4967360319923898e-05,
+      "loss": 3.8642,
+      "step": 2139648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4958990751217108e-05,
+      "loss": 3.8757,
+      "step": 2140160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4950604803706588e-05,
+      "loss": 3.859,
+      "step": 2140672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4942218856196068e-05,
+      "loss": 3.8569,
+      "step": 2141184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.493383290868555e-05,
+      "loss": 3.8751,
+      "step": 2141696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.492544696117503e-05,
+      "loss": 3.8702,
+      "step": 2142208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.491706101366451e-05,
+      "loss": 3.8683,
+      "step": 2142720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.490869144495772e-05,
+      "loss": 3.8739,
+      "step": 2143232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.49003054974472e-05,
+      "loss": 3.8547,
+      "step": 2143744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.489191954993668e-05,
+      "loss": 3.8555,
+      "step": 2144256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.488353360242616e-05,
+      "loss": 3.8476,
+      "step": 2144768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4875147654915639e-05,
+      "loss": 3.8656,
+      "step": 2145280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4866778086208851e-05,
+      "loss": 3.867,
+      "step": 2145792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4858392138698333e-05,
+      "loss": 3.86,
+      "step": 2146304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4850006191187813e-05,
+      "loss": 3.8628,
+      "step": 2146816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.484162024367729e-05,
+      "loss": 3.8823,
+      "step": 2147328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4833234296166771e-05,
+      "loss": 3.8672,
+      "step": 2147840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4824864727459984e-05,
+      "loss": 3.8717,
+      "step": 2148352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4816478779949464e-05,
+      "loss": 3.8654,
+      "step": 2148864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4808092832438942e-05,
+      "loss": 3.8698,
+      "step": 2149376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4799706884928422e-05,
+      "loss": 3.864,
+      "step": 2149888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4791337316221634e-05,
+      "loss": 3.8603,
+      "step": 2150400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4782951368711112e-05,
+      "loss": 3.8529,
+      "step": 2150912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4774565421200592e-05,
+      "loss": 3.8644,
+      "step": 2151424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4766179473690072e-05,
+      "loss": 3.8549,
+      "step": 2151936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4757793526179552e-05,
+      "loss": 3.8649,
+      "step": 2152448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4749423957472763e-05,
+      "loss": 3.8625,
+      "step": 2152960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4741038009962243e-05,
+      "loss": 3.8711,
+      "step": 2153472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4732668441255457e-05,
+      "loss": 3.8707,
+      "step": 2153984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4724282493744934e-05,
+      "loss": 3.86,
+      "step": 2154496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4715912925038148e-05,
+      "loss": 3.869,
+      "step": 2155008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4707526977527628e-05,
+      "loss": 3.8702,
+      "step": 2155520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4699141030017108e-05,
+      "loss": 3.8459,
+      "step": 2156032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4690755082506586e-05,
+      "loss": 3.8646,
+      "step": 2156544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4682369134996066e-05,
+      "loss": 3.8551,
+      "step": 2157056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4673983187485546e-05,
+      "loss": 3.8561,
+      "step": 2157568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4665597239975026e-05,
+      "loss": 3.8692,
+      "step": 2158080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4657227671268237e-05,
+      "loss": 3.8633,
+      "step": 2158592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4648841723757717e-05,
+      "loss": 3.8633,
+      "step": 2159104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4640455776247197e-05,
+      "loss": 3.8682,
+      "step": 2159616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4632069828736677e-05,
+      "loss": 3.8688,
+      "step": 2160128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4623683881226158e-05,
+      "loss": 3.8688,
+      "step": 2160640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4615314312519367e-05,
+      "loss": 3.8648,
+      "step": 2161152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4606928365008849e-05,
+      "loss": 3.8361,
+      "step": 2161664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4598542417498329e-05,
+      "loss": 3.8696,
+      "step": 2162176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4590156469987809e-05,
+      "loss": 3.8668,
+      "step": 2162688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4581770522477289e-05,
+      "loss": 3.8717,
+      "step": 2163200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4573384574966767e-05,
+      "loss": 3.8642,
+      "step": 2163712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4564998627456247e-05,
+      "loss": 3.8558,
+      "step": 2164224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4556612679945727e-05,
+      "loss": 3.8485,
+      "step": 2164736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4548243111238941e-05,
+      "loss": 3.8528,
+      "step": 2165248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4539857163728418e-05,
+      "loss": 3.8582,
+      "step": 2165760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4531471216217898e-05,
+      "loss": 3.8442,
+      "step": 2166272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.452308526870738e-05,
+      "loss": 3.8797,
+      "step": 2166784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.451469932119686e-05,
+      "loss": 3.8628,
+      "step": 2167296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.450631337368634e-05,
+      "loss": 3.8414,
+      "step": 2167808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.449792742617582e-05,
+      "loss": 3.8668,
+      "step": 2168320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4489541478665298e-05,
+      "loss": 3.8523,
+      "step": 2168832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4481155531154778e-05,
+      "loss": 3.8496,
+      "step": 2169344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.447278596244799e-05,
+      "loss": 3.861,
+      "step": 2169856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4464400014937472e-05,
+      "loss": 3.8619,
+      "step": 2170368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.445603044623068e-05,
+      "loss": 3.8479,
+      "step": 2170880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.444764449872016e-05,
+      "loss": 3.8613,
+      "step": 2171392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4439258551209642e-05,
+      "loss": 3.8483,
+      "step": 2171904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4430872603699122e-05,
+      "loss": 3.8566,
+      "step": 2172416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4422486656188599e-05,
+      "loss": 3.8658,
+      "step": 2172928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.441410070867808e-05,
+      "loss": 3.8581,
+      "step": 2173440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.440571476116756e-05,
+      "loss": 3.8618,
+      "step": 2173952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.439732881365704e-05,
+      "loss": 3.8584,
+      "step": 2174464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4388959244950251e-05,
+      "loss": 3.8815,
+      "step": 2174976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4380573297439731e-05,
+      "loss": 3.8449,
+      "step": 2175488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4372187349929211e-05,
+      "loss": 3.8691,
+      "step": 2176000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4363801402418691e-05,
+      "loss": 3.8552,
+      "step": 2176512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4355431833711902e-05,
+      "loss": 3.8582,
+      "step": 2177024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4347045886201382e-05,
+      "loss": 3.8598,
+      "step": 2177536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4338659938690862e-05,
+      "loss": 3.8601,
+      "step": 2178048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4330273991180344e-05,
+      "loss": 3.8757,
+      "step": 2178560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4321888043669823e-05,
+      "loss": 3.8705,
+      "step": 2179072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4313502096159303e-05,
+      "loss": 3.8447,
+      "step": 2179584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4305116148648782e-05,
+      "loss": 3.8454,
+      "step": 2180096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4296746579941994e-05,
+      "loss": 3.8507,
+      "step": 2180608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4288360632431474e-05,
+      "loss": 3.8671,
+      "step": 2181120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4279974684920952e-05,
+      "loss": 3.857,
+      "step": 2181632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4271588737410432e-05,
+      "loss": 3.8552,
+      "step": 2182144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4263219168703645e-05,
+      "loss": 3.8642,
+      "step": 2182656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4254833221193126e-05,
+      "loss": 3.8584,
+      "step": 2183168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4246447273682603e-05,
+      "loss": 3.8615,
+      "step": 2183680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4238061326172083e-05,
+      "loss": 3.8533,
+      "step": 2184192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4229675378661565e-05,
+      "loss": 3.8628,
+      "step": 2184704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4221289431151045e-05,
+      "loss": 3.8657,
+      "step": 2185216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4212903483640525e-05,
+      "loss": 3.8705,
+      "step": 2185728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4204533914933735e-05,
+      "loss": 3.8486,
+      "step": 2186240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4196147967423215e-05,
+      "loss": 3.8499,
+      "step": 2186752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4187762019912695e-05,
+      "loss": 3.8599,
+      "step": 2187264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4179376072402175e-05,
+      "loss": 3.8587,
+      "step": 2187776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4170990124891655e-05,
+      "loss": 3.8709,
+      "step": 2188288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4162620556184866e-05,
+      "loss": 3.8555,
+      "step": 2188800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4154234608674346e-05,
+      "loss": 3.8574,
+      "step": 2189312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4145848661163828e-05,
+      "loss": 3.8547,
+      "step": 2189824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4137462713653308e-05,
+      "loss": 3.8502,
+      "step": 2190336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4129076766142784e-05,
+      "loss": 3.8571,
+      "step": 2190848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4120707197435998e-05,
+      "loss": 3.8538,
+      "step": 2191360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4112321249925478e-05,
+      "loss": 3.862,
+      "step": 2191872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4103935302414956e-05,
+      "loss": 3.8557,
+      "step": 2192384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4095549354904436e-05,
+      "loss": 3.8546,
+      "step": 2192896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4087163407393916e-05,
+      "loss": 3.8564,
+      "step": 2193408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4078777459883396e-05,
+      "loss": 3.8457,
+      "step": 2193920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4070391512372876e-05,
+      "loss": 3.8639,
+      "step": 2194432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4062005564862358e-05,
+      "loss": 3.8622,
+      "step": 2194944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4053635996155567e-05,
+      "loss": 3.8642,
+      "step": 2195456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4045266427448781e-05,
+      "loss": 3.8673,
+      "step": 2195968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4036880479938258e-05,
+      "loss": 3.8654,
+      "step": 2196480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4028494532427738e-05,
+      "loss": 3.8579,
+      "step": 2196992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.402010858491722e-05,
+      "loss": 3.8513,
+      "step": 2197504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.40117226374067e-05,
+      "loss": 3.8648,
+      "step": 2198016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.400333668989618e-05,
+      "loss": 3.8576,
+      "step": 2198528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.399496712118939e-05,
+      "loss": 3.8579,
+      "step": 2199040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.398658117367887e-05,
+      "loss": 3.8633,
+      "step": 2199552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.397819522616835e-05,
+      "loss": 3.8519,
+      "step": 2200064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.396980927865783e-05,
+      "loss": 3.8614,
+      "step": 2200576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3961423331147312e-05,
+      "loss": 3.8503,
+      "step": 2201088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.395305376244052e-05,
+      "loss": 3.8564,
+      "step": 2201600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.394466781493e-05,
+      "loss": 3.8585,
+      "step": 2202112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3936281867419482e-05,
+      "loss": 3.8523,
+      "step": 2202624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3927895919908962e-05,
+      "loss": 3.8715,
+      "step": 2203136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3919526351202173e-05,
+      "loss": 3.8617,
+      "step": 2203648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3911140403691653e-05,
+      "loss": 3.863,
+      "step": 2204160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3902770834984862e-05,
+      "loss": 3.8528,
+      "step": 2204672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3894384887474344e-05,
+      "loss": 3.8594,
+      "step": 2205184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3885998939963824e-05,
+      "loss": 3.8642,
+      "step": 2205696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3877629371257035e-05,
+      "loss": 3.8648,
+      "step": 2206208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3869243423746514e-05,
+      "loss": 3.859,
+      "step": 2206720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3860857476235994e-05,
+      "loss": 3.8543,
+      "step": 2207232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3852471528725474e-05,
+      "loss": 3.8631,
+      "step": 2207744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3844085581214954e-05,
+      "loss": 3.8777,
+      "step": 2208256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3835699633704436e-05,
+      "loss": 3.8577,
+      "step": 2208768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3827313686193913e-05,
+      "loss": 3.8665,
+      "step": 2209280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3818927738683393e-05,
+      "loss": 3.869,
+      "step": 2209792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3810541791172874e-05,
+      "loss": 3.8615,
+      "step": 2210304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3802172222466083e-05,
+      "loss": 3.8632,
+      "step": 2210816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3793786274955565e-05,
+      "loss": 3.8579,
+      "step": 2211328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3785400327445045e-05,
+      "loss": 3.8607,
+      "step": 2211840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3777014379934525e-05,
+      "loss": 3.8633,
+      "step": 2212352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3768628432424005e-05,
+      "loss": 3.8544,
+      "step": 2212864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.984246015548706,
+      "eval_runtime": 286.454,
+      "eval_samples_per_second": 1332.12,
+      "eval_steps_per_second": 41.63,
+      "step": 2213280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3760242484913485e-05,
+      "loss": 3.8764,
+      "step": 2213376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3751856537402966e-05,
+      "loss": 3.8454,
+      "step": 2213888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3743470589892443e-05,
+      "loss": 3.8604,
+      "step": 2214400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3735084642381923e-05,
+      "loss": 3.8749,
+      "step": 2214912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3726698694871405e-05,
+      "loss": 3.8671,
+      "step": 2215424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3718312747360885e-05,
+      "loss": 3.86,
+      "step": 2215936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3709926799850364e-05,
+      "loss": 3.866,
+      "step": 2216448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3701540852339844e-05,
+      "loss": 3.8584,
+      "step": 2216960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3693154904829324e-05,
+      "loss": 3.851,
+      "step": 2217472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3684768957318803e-05,
+      "loss": 3.8725,
+      "step": 2217984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3676383009808283e-05,
+      "loss": 3.8609,
+      "step": 2218496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3667997062297763e-05,
+      "loss": 3.8654,
+      "step": 2219008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3659627493590973e-05,
+      "loss": 3.8688,
+      "step": 2219520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3651241546080453e-05,
+      "loss": 3.8499,
+      "step": 2220032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3642855598569935e-05,
+      "loss": 3.8519,
+      "step": 2220544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3634469651059415e-05,
+      "loss": 3.8459,
+      "step": 2221056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3626083703548895e-05,
+      "loss": 3.8631,
+      "step": 2221568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3617697756038375e-05,
+      "loss": 3.8608,
+      "step": 2222080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3609311808527855e-05,
+      "loss": 3.8573,
+      "step": 2222592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3600925861017333e-05,
+      "loss": 3.8557,
+      "step": 2223104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3592539913506813e-05,
+      "loss": 3.8798,
+      "step": 2223616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3584170344800025e-05,
+      "loss": 3.8614,
+      "step": 2224128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3575784397289504e-05,
+      "loss": 3.8666,
+      "step": 2224640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3567398449778984e-05,
+      "loss": 3.8637,
+      "step": 2225152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3559012502268464e-05,
+      "loss": 3.8675,
+      "step": 2225664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3550626554757945e-05,
+      "loss": 3.8579,
+      "step": 2226176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3542240607247425e-05,
+      "loss": 3.857,
+      "step": 2226688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3533854659736905e-05,
+      "loss": 3.8568,
+      "step": 2227200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3525468712226385e-05,
+      "loss": 3.8548,
+      "step": 2227712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3517099143519596e-05,
+      "loss": 3.8529,
+      "step": 2228224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3508713196009076e-05,
+      "loss": 3.8605,
+      "step": 2228736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3500343627302287e-05,
+      "loss": 3.8558,
+      "step": 2229248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3491957679791767e-05,
+      "loss": 3.871,
+      "step": 2229760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3483571732281247e-05,
+      "loss": 3.8632,
+      "step": 2230272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3475185784770728e-05,
+      "loss": 3.8583,
+      "step": 2230784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3466816216063937e-05,
+      "loss": 3.8677,
+      "step": 2231296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3458430268553417e-05,
+      "loss": 3.8644,
+      "step": 2231808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3450044321042899e-05,
+      "loss": 3.8397,
+      "step": 2232320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3441658373532379e-05,
+      "loss": 3.8616,
+      "step": 2232832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3433272426021859e-05,
+      "loss": 3.8507,
+      "step": 2233344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3424886478511337e-05,
+      "loss": 3.8534,
+      "step": 2233856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3416500531000817e-05,
+      "loss": 3.8634,
+      "step": 2234368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3408114583490297e-05,
+      "loss": 3.8631,
+      "step": 2234880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.339974501478351e-05,
+      "loss": 3.856,
+      "step": 2235392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.339137544607672e-05,
+      "loss": 3.8669,
+      "step": 2235904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.33829894985662e-05,
+      "loss": 3.8644,
+      "step": 2236416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3374603551055682e-05,
+      "loss": 3.8646,
+      "step": 2236928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3366217603545158e-05,
+      "loss": 3.8615,
+      "step": 2237440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3357831656034638e-05,
+      "loss": 3.8365,
+      "step": 2237952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3349445708524118e-05,
+      "loss": 3.8619,
+      "step": 2238464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.33410597610136e-05,
+      "loss": 3.8686,
+      "step": 2238976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.333267381350308e-05,
+      "loss": 3.8639,
+      "step": 2239488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3324320623600023e-05,
+      "loss": 3.8583,
+      "step": 2240000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3315934676089503e-05,
+      "loss": 3.8521,
+      "step": 2240512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3307548728578983e-05,
+      "loss": 3.8461,
+      "step": 2241024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3299162781068462e-05,
+      "loss": 3.8438,
+      "step": 2241536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3290776833557941e-05,
+      "loss": 3.8586,
+      "step": 2242048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3282390886047421e-05,
+      "loss": 3.8422,
+      "step": 2242560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3274004938536901e-05,
+      "loss": 3.8685,
+      "step": 2243072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3265618991026383e-05,
+      "loss": 3.8617,
+      "step": 2243584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3257249422319592e-05,
+      "loss": 3.8422,
+      "step": 2244096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3248879853612806e-05,
+      "loss": 3.8618,
+      "step": 2244608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3240493906102283e-05,
+      "loss": 3.8464,
+      "step": 2245120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3232107958591763e-05,
+      "loss": 3.8476,
+      "step": 2245632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3223722011081244e-05,
+      "loss": 3.8527,
+      "step": 2246144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3215336063570724e-05,
+      "loss": 3.8598,
+      "step": 2246656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3206950116060204e-05,
+      "loss": 3.8471,
+      "step": 2247168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3198580547353415e-05,
+      "loss": 3.8578,
+      "step": 2247680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3190194599842895e-05,
+      "loss": 3.843,
+      "step": 2248192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3181808652332375e-05,
+      "loss": 3.8513,
+      "step": 2248704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3173422704821855e-05,
+      "loss": 3.8595,
+      "step": 2249216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3165036757311337e-05,
+      "loss": 3.8575,
+      "step": 2249728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3156667188604546e-05,
+      "loss": 3.8552,
+      "step": 2250240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3148281241094026e-05,
+      "loss": 3.8559,
+      "step": 2250752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3139895293583507e-05,
+      "loss": 3.8744,
+      "step": 2251264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3131509346072987e-05,
+      "loss": 3.8415,
+      "step": 2251776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3123123398562464e-05,
+      "loss": 3.8656,
+      "step": 2252288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3114753829855678e-05,
+      "loss": 3.8526,
+      "step": 2252800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3106367882345158e-05,
+      "loss": 3.8514,
+      "step": 2253312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3097981934834638e-05,
+      "loss": 3.8534,
+      "step": 2253824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3089595987324116e-05,
+      "loss": 3.8572,
+      "step": 2254336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3081210039813596e-05,
+      "loss": 3.8685,
+      "step": 2254848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3072824092303076e-05,
+      "loss": 3.8698,
+      "step": 2255360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3064438144792556e-05,
+      "loss": 3.8446,
+      "step": 2255872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3056068576085767e-05,
+      "loss": 3.8424,
+      "step": 2256384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3047682628575247e-05,
+      "loss": 3.8475,
+      "step": 2256896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3039296681064729e-05,
+      "loss": 3.8608,
+      "step": 2257408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3030910733554208e-05,
+      "loss": 3.8526,
+      "step": 2257920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3022524786043688e-05,
+      "loss": 3.8501,
+      "step": 2258432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3014171596140632e-05,
+      "loss": 3.8587,
+      "step": 2258944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3005785648630112e-05,
+      "loss": 3.8532,
+      "step": 2259456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.299739970111959e-05,
+      "loss": 3.8618,
+      "step": 2259968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.298901375360907e-05,
+      "loss": 3.8468,
+      "step": 2260480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2980644184902282e-05,
+      "loss": 3.8591,
+      "step": 2260992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.297225823739176e-05,
+      "loss": 3.8549,
+      "step": 2261504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.296387228988124e-05,
+      "loss": 3.869,
+      "step": 2262016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.295548634237072e-05,
+      "loss": 3.8425,
+      "step": 2262528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.29471003948602e-05,
+      "loss": 3.8469,
+      "step": 2263040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.293871444734968e-05,
+      "loss": 3.8602,
+      "step": 2263552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2930328499839162e-05,
+      "loss": 3.8537,
+      "step": 2264064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2921942552328642e-05,
+      "loss": 3.8669,
+      "step": 2264576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2913556604818119e-05,
+      "loss": 3.8512,
+      "step": 2265088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.29051706573076e-05,
+      "loss": 3.8527,
+      "step": 2265600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.289678470979708e-05,
+      "loss": 3.8512,
+      "step": 2266112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.288839876228656e-05,
+      "loss": 3.8502,
+      "step": 2266624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.288001281477604e-05,
+      "loss": 3.8501,
+      "step": 2267136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2871626867265522e-05,
+      "loss": 3.848,
+      "step": 2267648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2863240919754998e-05,
+      "loss": 3.8566,
+      "step": 2268160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2854854972244478e-05,
+      "loss": 3.8492,
+      "step": 2268672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2846485403537693e-05,
+      "loss": 3.8503,
+      "step": 2269184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2838099456027172e-05,
+      "loss": 3.8529,
+      "step": 2269696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2829713508516649e-05,
+      "loss": 3.8412,
+      "step": 2270208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.282132756100613e-05,
+      "loss": 3.8643,
+      "step": 2270720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2812957992299343e-05,
+      "loss": 3.8563,
+      "step": 2271232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2804572044788823e-05,
+      "loss": 3.8561,
+      "step": 2271744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2796186097278301e-05,
+      "loss": 3.864,
+      "step": 2272256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2787800149767781e-05,
+      "loss": 3.8621,
+      "step": 2272768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2779446959864725e-05,
+      "loss": 3.8519,
+      "step": 2273280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2771077391157937e-05,
+      "loss": 3.848,
+      "step": 2273792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2762691443647415e-05,
+      "loss": 3.8588,
+      "step": 2274304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2754305496136895e-05,
+      "loss": 3.8538,
+      "step": 2274816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2745919548626375e-05,
+      "loss": 3.8508,
+      "step": 2275328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2737533601115855e-05,
+      "loss": 3.862,
+      "step": 2275840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2729147653605337e-05,
+      "loss": 3.8535,
+      "step": 2276352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2720761706094817e-05,
+      "loss": 3.8561,
+      "step": 2276864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2712392137388026e-05,
+      "loss": 3.8477,
+      "step": 2277376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2704006189877508e-05,
+      "loss": 3.8494,
+      "step": 2277888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2695636621170717e-05,
+      "loss": 3.8569,
+      "step": 2278400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2687250673660198e-05,
+      "loss": 3.8489,
+      "step": 2278912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2678864726149678e-05,
+      "loss": 3.8601,
+      "step": 2279424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2670478778639158e-05,
+      "loss": 3.8671,
+      "step": 2279936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2662092831128638e-05,
+      "loss": 3.856,
+      "step": 2280448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2653706883618118e-05,
+      "loss": 3.8485,
+      "step": 2280960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2645320936107596e-05,
+      "loss": 3.8619,
+      "step": 2281472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2636934988597076e-05,
+      "loss": 3.8581,
+      "step": 2281984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2628549041086556e-05,
+      "loss": 3.8579,
+      "step": 2282496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.262017947237977e-05,
+      "loss": 3.8508,
+      "step": 2283008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2611793524869247e-05,
+      "loss": 3.8515,
+      "step": 2283520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2603407577358727e-05,
+      "loss": 3.8647,
+      "step": 2284032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2595021629848209e-05,
+      "loss": 3.8719,
+      "step": 2284544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2586635682337689e-05,
+      "loss": 3.8501,
+      "step": 2285056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.25782661136309e-05,
+      "loss": 3.8626,
+      "step": 2285568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.256988016612038e-05,
+      "loss": 3.8644,
+      "step": 2286080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.256149421860986e-05,
+      "loss": 3.8562,
+      "step": 2286592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.255312464990307e-05,
+      "loss": 3.8594,
+      "step": 2287104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.254473870239255e-05,
+      "loss": 3.8526,
+      "step": 2287616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.253635275488203e-05,
+      "loss": 3.8561,
+      "step": 2288128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.252796680737151e-05,
+      "loss": 3.8649,
+      "step": 2288640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2519580859860992e-05,
+      "loss": 3.8438,
+      "step": 2289152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.982982873916626,
+      "eval_runtime": 291.1868,
+      "eval_samples_per_second": 1310.468,
+      "eval_steps_per_second": 40.953,
+      "step": 2289600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2511194912350472e-05,
+      "loss": 3.8559,
+      "step": 2289664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2502808964839952e-05,
+      "loss": 3.8451,
+      "step": 2290176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.249442301732943e-05,
+      "loss": 3.8521,
+      "step": 2290688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.248603706981891e-05,
+      "loss": 3.8676,
+      "step": 2291200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.247765112230839e-05,
+      "loss": 3.8659,
+      "step": 2291712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.246926517479787e-05,
+      "loss": 3.8552,
+      "step": 2292224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.246087922728735e-05,
+      "loss": 3.8626,
+      "step": 2292736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.245250965858056e-05,
+      "loss": 3.8511,
+      "step": 2293248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2444140089873773e-05,
+      "loss": 3.8472,
+      "step": 2293760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2435754142363253e-05,
+      "loss": 3.8673,
+      "step": 2294272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2427368194852733e-05,
+      "loss": 3.86,
+      "step": 2294784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2418998626145944e-05,
+      "loss": 3.8629,
+      "step": 2295296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2410629057439154e-05,
+      "loss": 3.8662,
+      "step": 2295808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2402243109928634e-05,
+      "loss": 3.8464,
+      "step": 2296320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2393857162418116e-05,
+      "loss": 3.8469,
+      "step": 2296832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2385471214907594e-05,
+      "loss": 3.8397,
+      "step": 2297344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2377085267397074e-05,
+      "loss": 3.8591,
+      "step": 2297856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2368699319886554e-05,
+      "loss": 3.8546,
+      "step": 2298368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2360313372376034e-05,
+      "loss": 3.8541,
+      "step": 2298880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2351927424865514e-05,
+      "loss": 3.8552,
+      "step": 2299392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2343541477354994e-05,
+      "loss": 3.8729,
+      "step": 2299904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2335155529844474e-05,
+      "loss": 3.8635,
+      "step": 2300416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2326769582333954e-05,
+      "loss": 3.8559,
+      "step": 2300928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2318383634823434e-05,
+      "loss": 3.8589,
+      "step": 2301440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2309997687312912e-05,
+      "loss": 3.863,
+      "step": 2301952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2301611739802394e-05,
+      "loss": 3.8551,
+      "step": 2302464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2293225792291874e-05,
+      "loss": 3.849,
+      "step": 2302976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2284839844781354e-05,
+      "loss": 3.8536,
+      "step": 2303488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2276470276074565e-05,
+      "loss": 3.8485,
+      "step": 2304000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2268084328564045e-05,
+      "loss": 3.8531,
+      "step": 2304512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2259698381053525e-05,
+      "loss": 3.8557,
+      "step": 2305024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2251312433543004e-05,
+      "loss": 3.8549,
+      "step": 2305536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2242942864836215e-05,
+      "loss": 3.86,
+      "step": 2306048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2234556917325695e-05,
+      "loss": 3.862,
+      "step": 2306560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2226170969815177e-05,
+      "loss": 3.8543,
+      "step": 2307072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2217801401108386e-05,
+      "loss": 3.8633,
+      "step": 2307584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2209415453597866e-05,
+      "loss": 3.862,
+      "step": 2308096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2201029506087348e-05,
+      "loss": 3.8364,
+      "step": 2308608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2192643558576828e-05,
+      "loss": 3.8633,
+      "step": 2309120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2184257611066306e-05,
+      "loss": 3.8406,
+      "step": 2309632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2175871663555786e-05,
+      "loss": 3.8515,
+      "step": 2310144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2167485716045267e-05,
+      "loss": 3.8541,
+      "step": 2310656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2159099768534746e-05,
+      "loss": 3.8633,
+      "step": 2311168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2150730199827958e-05,
+      "loss": 3.85,
+      "step": 2311680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2142344252317438e-05,
+      "loss": 3.8598,
+      "step": 2312192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2133974683610649e-05,
+      "loss": 3.8645,
+      "step": 2312704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2125588736100129e-05,
+      "loss": 3.8568,
+      "step": 2313216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2117202788589609e-05,
+      "loss": 3.8618,
+      "step": 2313728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2108816841079089e-05,
+      "loss": 3.8297,
+      "step": 2314240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2100430893568567e-05,
+      "loss": 3.855,
+      "step": 2314752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.209206132486178e-05,
+      "loss": 3.8664,
+      "step": 2315264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.208367537735126e-05,
+      "loss": 3.8597,
+      "step": 2315776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.207528942984074e-05,
+      "loss": 3.8535,
+      "step": 2316288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.206690348233022e-05,
+      "loss": 3.8455,
+      "step": 2316800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.205853391362343e-05,
+      "loss": 3.8484,
+      "step": 2317312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2050147966112912e-05,
+      "loss": 3.836,
+      "step": 2317824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2041762018602392e-05,
+      "loss": 3.861,
+      "step": 2318336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.203337607109187e-05,
+      "loss": 3.8373,
+      "step": 2318848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2025006502385082e-05,
+      "loss": 3.8605,
+      "step": 2319360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2016620554874562e-05,
+      "loss": 3.8542,
+      "step": 2319872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.200823460736404e-05,
+      "loss": 3.8421,
+      "step": 2320384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.199984865985352e-05,
+      "loss": 3.8574,
+      "step": 2320896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1991462712343002e-05,
+      "loss": 3.8411,
+      "step": 2321408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1983076764832482e-05,
+      "loss": 3.8474,
+      "step": 2321920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.197469081732196e-05,
+      "loss": 3.8464,
+      "step": 2322432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.196630486981144e-05,
+      "loss": 3.8569,
+      "step": 2322944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1957935301104653e-05,
+      "loss": 3.8426,
+      "step": 2323456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1949549353594131e-05,
+      "loss": 3.8513,
+      "step": 2323968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1941163406083613e-05,
+      "loss": 3.8393,
+      "step": 2324480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1932777458573093e-05,
+      "loss": 3.8458,
+      "step": 2324992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1924391511062573e-05,
+      "loss": 3.861,
+      "step": 2325504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1916021942355784e-05,
+      "loss": 3.856,
+      "step": 2326016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1907635994845264e-05,
+      "loss": 3.8476,
+      "step": 2326528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1899250047334744e-05,
+      "loss": 3.8541,
+      "step": 2327040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1890864099824223e-05,
+      "loss": 3.8685,
+      "step": 2327552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1882494531117434e-05,
+      "loss": 3.8428,
+      "step": 2328064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1874108583606914e-05,
+      "loss": 3.859,
+      "step": 2328576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1865722636096394e-05,
+      "loss": 3.8486,
+      "step": 2329088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1857336688585874e-05,
+      "loss": 3.8519,
+      "step": 2329600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1848950741075354e-05,
+      "loss": 3.8455,
+      "step": 2330112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1840564793564834e-05,
+      "loss": 3.8526,
+      "step": 2330624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1832195224858047e-05,
+      "loss": 3.8655,
+      "step": 2331136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1823809277347525e-05,
+      "loss": 3.8617,
+      "step": 2331648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1815423329837005e-05,
+      "loss": 3.8479,
+      "step": 2332160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1807037382326486e-05,
+      "loss": 3.8328,
+      "step": 2332672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1798651434815965e-05,
+      "loss": 3.8412,
+      "step": 2333184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1790281866109177e-05,
+      "loss": 3.8579,
+      "step": 2333696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1781895918598657e-05,
+      "loss": 3.848,
+      "step": 2334208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1773509971088135e-05,
+      "loss": 3.8479,
+      "step": 2334720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1765124023577615e-05,
+      "loss": 3.8557,
+      "step": 2335232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1756738076067097e-05,
+      "loss": 3.8514,
+      "step": 2335744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1748368507360308e-05,
+      "loss": 3.8521,
+      "step": 2336256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1739982559849786e-05,
+      "loss": 3.8455,
+      "step": 2336768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1731596612339268e-05,
+      "loss": 3.8566,
+      "step": 2337280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1723210664828748e-05,
+      "loss": 3.848,
+      "step": 2337792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1714824717318226e-05,
+      "loss": 3.8648,
+      "step": 2338304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1706455148611438e-05,
+      "loss": 3.8421,
+      "step": 2338816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1698069201100918e-05,
+      "loss": 3.8421,
+      "step": 2339328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1689683253590398e-05,
+      "loss": 3.8569,
+      "step": 2339840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1681313684883609e-05,
+      "loss": 3.8506,
+      "step": 2340352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1672927737373089e-05,
+      "loss": 3.8587,
+      "step": 2340864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1664541789862569e-05,
+      "loss": 3.8502,
+      "step": 2341376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.165615584235205e-05,
+      "loss": 3.8508,
+      "step": 2341888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1647769894841529e-05,
+      "loss": 3.8453,
+      "step": 2342400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1639383947331009e-05,
+      "loss": 3.8503,
+      "step": 2342912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1630997999820489e-05,
+      "loss": 3.8423,
+      "step": 2343424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1622612052309969e-05,
+      "loss": 3.847,
+      "step": 2343936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1614226104799449e-05,
+      "loss": 3.8518,
+      "step": 2344448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1605840157288929e-05,
+      "loss": 3.8472,
+      "step": 2344960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1597454209778409e-05,
+      "loss": 3.8447,
+      "step": 2345472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1589068262267889e-05,
+      "loss": 3.8536,
+      "step": 2345984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.15806986935611e-05,
+      "loss": 3.8348,
+      "step": 2346496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.157231274605058e-05,
+      "loss": 3.8576,
+      "step": 2347008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.156392679854006e-05,
+      "loss": 3.8526,
+      "step": 2347520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.155554085102954e-05,
+      "loss": 3.8559,
+      "step": 2348032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1547171282322752e-05,
+      "loss": 3.859,
+      "step": 2348544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1538785334812232e-05,
+      "loss": 3.8592,
+      "step": 2349056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.153039938730171e-05,
+      "loss": 3.8502,
+      "step": 2349568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.152201343979119e-05,
+      "loss": 3.8471,
+      "step": 2350080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1513643871084402e-05,
+      "loss": 3.8527,
+      "step": 2350592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.150525792357388e-05,
+      "loss": 3.8523,
+      "step": 2351104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.149687197606336e-05,
+      "loss": 3.8424,
+      "step": 2351616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1488502407356573e-05,
+      "loss": 3.8618,
+      "step": 2352128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1480116459846053e-05,
+      "loss": 3.8469,
+      "step": 2352640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1471730512335533e-05,
+      "loss": 3.8545,
+      "step": 2353152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1463344564825013e-05,
+      "loss": 3.8439,
+      "step": 2353664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1454958617314493e-05,
+      "loss": 3.8423,
+      "step": 2354176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1446572669803971e-05,
+      "loss": 3.8552,
+      "step": 2354688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1438186722293453e-05,
+      "loss": 3.8489,
+      "step": 2355200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1429800774782933e-05,
+      "loss": 3.8472,
+      "step": 2355712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1421431206076144e-05,
+      "loss": 3.8647,
+      "step": 2356224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1413045258565624e-05,
+      "loss": 3.8502,
+      "step": 2356736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1404659311055103e-05,
+      "loss": 3.8455,
+      "step": 2357248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1396273363544583e-05,
+      "loss": 3.8575,
+      "step": 2357760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1387887416034063e-05,
+      "loss": 3.8563,
+      "step": 2358272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1379517847327274e-05,
+      "loss": 3.8556,
+      "step": 2358784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1371131899816754e-05,
+      "loss": 3.8443,
+      "step": 2359296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1362745952306234e-05,
+      "loss": 3.848,
+      "step": 2359808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1354376383599445e-05,
+      "loss": 3.8654,
+      "step": 2360320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1345990436088925e-05,
+      "loss": 3.8656,
+      "step": 2360832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1337604488578406e-05,
+      "loss": 3.8491,
+      "step": 2361344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1329218541067886e-05,
+      "loss": 3.8537,
+      "step": 2361856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1320832593557365e-05,
+      "loss": 3.861,
+      "step": 2362368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1312446646046845e-05,
+      "loss": 3.8543,
+      "step": 2362880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1304077077340057e-05,
+      "loss": 3.855,
+      "step": 2363392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1295691129829535e-05,
+      "loss": 3.8508,
+      "step": 2363904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1287305182319017e-05,
+      "loss": 3.8495,
+      "step": 2364416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1278919234808497e-05,
+      "loss": 3.8623,
+      "step": 2364928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.1270533287297977e-05,
+      "loss": 3.842,
+      "step": 2365440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9820263385772705,
+      "eval_runtime": 289.2479,
+      "eval_samples_per_second": 1319.252,
+      "eval_steps_per_second": 41.228,
+      "step": 2365920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1262147339787455e-05,
+      "loss": 3.828,
+      "step": 2365952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1253761392276937e-05,
+      "loss": 3.8455,
+      "step": 2366464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1245375444766417e-05,
+      "loss": 3.8465,
+      "step": 2366976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1236989497255895e-05,
+      "loss": 3.8656,
+      "step": 2367488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1228603549745375e-05,
+      "loss": 3.8639,
+      "step": 2368000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1220217602234857e-05,
+      "loss": 3.8508,
+      "step": 2368512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1211831654724335e-05,
+      "loss": 3.8564,
+      "step": 2369024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1203445707213815e-05,
+      "loss": 3.8515,
+      "step": 2369536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1195059759703295e-05,
+      "loss": 3.8495,
+      "step": 2370048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1186673812192775e-05,
+      "loss": 3.861,
+      "step": 2370560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1178287864682255e-05,
+      "loss": 3.8531,
+      "step": 2371072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1169901917171735e-05,
+      "loss": 3.8585,
+      "step": 2371584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1161532348464947e-05,
+      "loss": 3.8641,
+      "step": 2372096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1153146400954425e-05,
+      "loss": 3.8448,
+      "step": 2372608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1144760453443905e-05,
+      "loss": 3.8433,
+      "step": 2373120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1136374505933387e-05,
+      "loss": 3.8361,
+      "step": 2373632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1127988558422865e-05,
+      "loss": 3.8539,
+      "step": 2374144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1119602610912345e-05,
+      "loss": 3.8522,
+      "step": 2374656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1111216663401825e-05,
+      "loss": 3.8502,
+      "step": 2375168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1102830715891307e-05,
+      "loss": 3.8512,
+      "step": 2375680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1094444768380785e-05,
+      "loss": 3.8676,
+      "step": 2376192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1086075199673996e-05,
+      "loss": 3.8554,
+      "step": 2376704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1077689252163478e-05,
+      "loss": 3.8589,
+      "step": 2377216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1069303304652956e-05,
+      "loss": 3.8535,
+      "step": 2377728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1060917357142436e-05,
+      "loss": 3.8637,
+      "step": 2378240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1052531409631916e-05,
+      "loss": 3.8494,
+      "step": 2378752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1044161840925128e-05,
+      "loss": 3.8523,
+      "step": 2379264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1035775893414607e-05,
+      "loss": 3.8471,
+      "step": 2379776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1027389945904088e-05,
+      "loss": 3.8444,
+      "step": 2380288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1019003998393568e-05,
+      "loss": 3.8486,
+      "step": 2380800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1010618050883046e-05,
+      "loss": 3.8491,
+      "step": 2381312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1002232103372526e-05,
+      "loss": 3.8514,
+      "step": 2381824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0993846155862008e-05,
+      "loss": 3.858,
+      "step": 2382336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0985460208351488e-05,
+      "loss": 3.8554,
+      "step": 2382848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0977090639644699e-05,
+      "loss": 3.8589,
+      "step": 2383360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.096872107093791e-05,
+      "loss": 3.8543,
+      "step": 2383872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0960351502231122e-05,
+      "loss": 3.8589,
+      "step": 2384384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0951965554720602e-05,
+      "loss": 3.8354,
+      "step": 2384896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.094357960721008e-05,
+      "loss": 3.8593,
+      "step": 2385408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.093519365969956e-05,
+      "loss": 3.8342,
+      "step": 2385920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0926807712189042e-05,
+      "loss": 3.8484,
+      "step": 2386432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.091842176467852e-05,
+      "loss": 3.8523,
+      "step": 2386944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0910035817168e-05,
+      "loss": 3.8594,
+      "step": 2387456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.090164986965748e-05,
+      "loss": 3.8463,
+      "step": 2387968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0893263922146962e-05,
+      "loss": 3.8581,
+      "step": 2388480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.088487797463644e-05,
+      "loss": 3.8598,
+      "step": 2388992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.087649202712592e-05,
+      "loss": 3.8559,
+      "step": 2389504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0868122458419132e-05,
+      "loss": 3.8522,
+      "step": 2390016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.085973651090861e-05,
+      "loss": 3.8299,
+      "step": 2390528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.085135056339809e-05,
+      "loss": 3.849,
+      "step": 2391040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0842964615887572e-05,
+      "loss": 3.8625,
+      "step": 2391552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0834578668377052e-05,
+      "loss": 3.8553,
+      "step": 2392064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.082619272086653e-05,
+      "loss": 3.8483,
+      "step": 2392576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.081780677335601e-05,
+      "loss": 3.8426,
+      "step": 2393088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0809437204649223e-05,
+      "loss": 3.8435,
+      "step": 2393600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0801051257138701e-05,
+      "loss": 3.8331,
+      "step": 2394112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0792681688431914e-05,
+      "loss": 3.8607,
+      "step": 2394624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0784312119725124e-05,
+      "loss": 3.8318,
+      "step": 2395136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0775926172214604e-05,
+      "loss": 3.8544,
+      "step": 2395648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0767540224704084e-05,
+      "loss": 3.855,
+      "step": 2396160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0759154277193564e-05,
+      "loss": 3.8428,
+      "step": 2396672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0750768329683044e-05,
+      "loss": 3.8478,
+      "step": 2397184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0742382382172524e-05,
+      "loss": 3.8406,
+      "step": 2397696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0733996434662004e-05,
+      "loss": 3.8434,
+      "step": 2398208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0725610487151484e-05,
+      "loss": 3.8376,
+      "step": 2398720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0717224539640964e-05,
+      "loss": 3.857,
+      "step": 2399232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0708838592130444e-05,
+      "loss": 3.8437,
+      "step": 2399744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0700452644619924e-05,
+      "loss": 3.8411,
+      "step": 2400256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0692083075913135e-05,
+      "loss": 3.8388,
+      "step": 2400768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0683697128402616e-05,
+      "loss": 3.8445,
+      "step": 2401280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0675311180892095e-05,
+      "loss": 3.8592,
+      "step": 2401792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0666925233381575e-05,
+      "loss": 3.8509,
+      "step": 2402304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0658539285871055e-05,
+      "loss": 3.8434,
+      "step": 2402816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0650153338360535e-05,
+      "loss": 3.8534,
+      "step": 2403328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0641783769653745e-05,
+      "loss": 3.8633,
+      "step": 2403840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0633397822143227e-05,
+      "loss": 3.8364,
+      "step": 2404352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0625028253436438e-05,
+      "loss": 3.8568,
+      "step": 2404864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0616642305925916e-05,
+      "loss": 3.8455,
+      "step": 2405376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0608256358415398e-05,
+      "loss": 3.8475,
+      "step": 2405888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0599870410904878e-05,
+      "loss": 3.8421,
+      "step": 2406400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0591484463394356e-05,
+      "loss": 3.8503,
+      "step": 2406912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0583098515883836e-05,
+      "loss": 3.86,
+      "step": 2407424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0574712568373318e-05,
+      "loss": 3.8583,
+      "step": 2407936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0566326620862796e-05,
+      "loss": 3.8476,
+      "step": 2408448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0557940673352276e-05,
+      "loss": 3.8244,
+      "step": 2408960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0549554725841757e-05,
+      "loss": 3.8439,
+      "step": 2409472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0541168778331237e-05,
+      "loss": 3.8464,
+      "step": 2409984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0532799209624446e-05,
+      "loss": 3.8514,
+      "step": 2410496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0524413262113928e-05,
+      "loss": 3.8393,
+      "step": 2411008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0516027314603408e-05,
+      "loss": 3.8547,
+      "step": 2411520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0507641367092886e-05,
+      "loss": 3.8492,
+      "step": 2412032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0499255419582366e-05,
+      "loss": 3.8511,
+      "step": 2412544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0490885850875579e-05,
+      "loss": 3.8399,
+      "step": 2413056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0482499903365059e-05,
+      "loss": 3.8547,
+      "step": 2413568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0474113955854539e-05,
+      "loss": 3.843,
+      "step": 2414080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0465728008344019e-05,
+      "loss": 3.8627,
+      "step": 2414592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0457342060833499e-05,
+      "loss": 3.846,
+      "step": 2415104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.044897249212671e-05,
+      "loss": 3.8365,
+      "step": 2415616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.044058654461619e-05,
+      "loss": 3.8512,
+      "step": 2416128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.043220059710567e-05,
+      "loss": 3.8501,
+      "step": 2416640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.042381464959515e-05,
+      "loss": 3.8553,
+      "step": 2417152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.041542870208463e-05,
+      "loss": 3.8475,
+      "step": 2417664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0407042754574109e-05,
+      "loss": 3.8451,
+      "step": 2418176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0398656807063589e-05,
+      "loss": 3.8413,
+      "step": 2418688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0390287238356802e-05,
+      "loss": 3.8486,
+      "step": 2419200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.038190129084628e-05,
+      "loss": 3.8362,
+      "step": 2419712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.037351534333576e-05,
+      "loss": 3.8471,
+      "step": 2420224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0365145774628972e-05,
+      "loss": 3.8473,
+      "step": 2420736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.035675982711845e-05,
+      "loss": 3.8478,
+      "step": 2421248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.034837387960793e-05,
+      "loss": 3.837,
+      "step": 2421760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0339987932097412e-05,
+      "loss": 3.8516,
+      "step": 2422272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0331601984586892e-05,
+      "loss": 3.8376,
+      "step": 2422784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.032321603707637e-05,
+      "loss": 3.8527,
+      "step": 2423296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.031483008956585e-05,
+      "loss": 3.8467,
+      "step": 2423808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0306444142055332e-05,
+      "loss": 3.8474,
+      "step": 2424320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.029805819454481e-05,
+      "loss": 3.8554,
+      "step": 2424832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.028967224703429e-05,
+      "loss": 3.856,
+      "step": 2425344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.028128629952377e-05,
+      "loss": 3.8506,
+      "step": 2425856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.027290035201325e-05,
+      "loss": 3.84,
+      "step": 2426368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0264530783306461e-05,
+      "loss": 3.8514,
+      "step": 2426880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.025614483579594e-05,
+      "loss": 3.8495,
+      "step": 2427392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0247758888285422e-05,
+      "loss": 3.8402,
+      "step": 2427904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0239389319578632e-05,
+      "loss": 3.8554,
+      "step": 2428416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0231003372068113e-05,
+      "loss": 3.8443,
+      "step": 2428928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0222617424557593e-05,
+      "loss": 3.8522,
+      "step": 2429440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0214231477047073e-05,
+      "loss": 3.8391,
+      "step": 2429952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0205845529536551e-05,
+      "loss": 3.8424,
+      "step": 2430464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0197459582026033e-05,
+      "loss": 3.8488,
+      "step": 2430976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0189073634515513e-05,
+      "loss": 3.8445,
+      "step": 2431488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0180687687004991e-05,
+      "loss": 3.84,
+      "step": 2432000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0172318118298204e-05,
+      "loss": 3.8633,
+      "step": 2432512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0163932170787684e-05,
+      "loss": 3.8506,
+      "step": 2433024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0155546223277164e-05,
+      "loss": 3.8427,
+      "step": 2433536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0147160275766644e-05,
+      "loss": 3.8513,
+      "step": 2434048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0138774328256124e-05,
+      "loss": 3.8533,
+      "step": 2434560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0130388380745604e-05,
+      "loss": 3.8514,
+      "step": 2435072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0122002433235082e-05,
+      "loss": 3.8463,
+      "step": 2435584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0113632864528294e-05,
+      "loss": 3.8436,
+      "step": 2436096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0105263295821505e-05,
+      "loss": 3.8605,
+      "step": 2436608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0096877348310987e-05,
+      "loss": 3.8616,
+      "step": 2437120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0088491400800465e-05,
+      "loss": 3.8482,
+      "step": 2437632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0080105453289945e-05,
+      "loss": 3.8506,
+      "step": 2438144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0071719505779425e-05,
+      "loss": 3.8597,
+      "step": 2438656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0063333558268905e-05,
+      "loss": 3.8485,
+      "step": 2439168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0054963989562116e-05,
+      "loss": 3.8515,
+      "step": 2439680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0046578042051597e-05,
+      "loss": 3.8489,
+      "step": 2440192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0038192094541077e-05,
+      "loss": 3.8462,
+      "step": 2440704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0029806147030555e-05,
+      "loss": 3.859,
+      "step": 2441216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0021420199520035e-05,
+      "loss": 3.8419,
+      "step": 2441728
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0013034252009517e-05,
+      "loss": 3.8454,
+      "step": 2442240
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.980874538421631,
+      "eval_runtime": 285.2859,
+      "eval_samples_per_second": 1337.574,
+      "eval_steps_per_second": 41.8,
+      "step": 2442240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0004648304498995e-05,
+      "loss": 3.8405,
+      "step": 2442752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.996262356988475e-06,
+      "loss": 3.8422,
+      "step": 2443264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.987876409477955e-06,
+      "loss": 3.8612,
+      "step": 2443776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.979506840771168e-06,
+      "loss": 3.8625,
+      "step": 2444288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.971120893260646e-06,
+      "loss": 3.8465,
+      "step": 2444800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.962734945750126e-06,
+      "loss": 3.8536,
+      "step": 2445312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.954365377043338e-06,
+      "loss": 3.849,
+      "step": 2445824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.945979429532817e-06,
+      "loss": 3.8393,
+      "step": 2446336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.937593482022298e-06,
+      "loss": 3.8593,
+      "step": 2446848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.929207534511778e-06,
+      "loss": 3.8492,
+      "step": 2447360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.920821587001258e-06,
+      "loss": 3.8573,
+      "step": 2447872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.912435639490737e-06,
+      "loss": 3.8558,
+      "step": 2448384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.904066070783949e-06,
+      "loss": 3.8491,
+      "step": 2448896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.895680123273429e-06,
+      "loss": 3.8397,
+      "step": 2449408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.887294175762909e-06,
+      "loss": 3.831,
+      "step": 2449920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.878908228252389e-06,
+      "loss": 3.8477,
+      "step": 2450432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.870522280741869e-06,
+      "loss": 3.8484,
+      "step": 2450944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.862136333231349e-06,
+      "loss": 3.8458,
+      "step": 2451456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.853750385720829e-06,
+      "loss": 3.8457,
+      "step": 2451968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.845364438210309e-06,
+      "loss": 3.8681,
+      "step": 2452480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.83699486950352e-06,
+      "loss": 3.8501,
+      "step": 2452992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.828608921993e-06,
+      "loss": 3.8534,
+      "step": 2453504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.82022297448248e-06,
+      "loss": 3.849,
+      "step": 2454016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.81183702697196e-06,
+      "loss": 3.8633,
+      "step": 2454528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.80345107946144e-06,
+      "loss": 3.841,
+      "step": 2455040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.79506513195092e-06,
+      "loss": 3.8474,
+      "step": 2455552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.7866791844404e-06,
+      "loss": 3.8434,
+      "step": 2456064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.77829323692988e-06,
+      "loss": 3.8446,
+      "step": 2456576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.76992366822309e-06,
+      "loss": 3.8453,
+      "step": 2457088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.76153772071257e-06,
+      "loss": 3.8446,
+      "step": 2457600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.753168152005782e-06,
+      "loss": 3.8487,
+      "step": 2458112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.744798583298993e-06,
+      "loss": 3.8546,
+      "step": 2458624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.736412635788471e-06,
+      "loss": 3.8487,
+      "step": 2459136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.728026688277953e-06,
+      "loss": 3.8561,
+      "step": 2459648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.719657119571164e-06,
+      "loss": 3.8499,
+      "step": 2460160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.711271172060644e-06,
+      "loss": 3.8556,
+      "step": 2460672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.702885224550124e-06,
+      "loss": 3.8331,
+      "step": 2461184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.694499277039604e-06,
+      "loss": 3.8558,
+      "step": 2461696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.686113329529084e-06,
+      "loss": 3.8301,
+      "step": 2462208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.677727382018564e-06,
+      "loss": 3.8483,
+      "step": 2462720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.669341434508044e-06,
+      "loss": 3.8446,
+      "step": 2463232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.660955486997524e-06,
+      "loss": 3.8575,
+      "step": 2463744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.652569539487004e-06,
+      "loss": 3.8409,
+      "step": 2464256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.644183591976484e-06,
+      "loss": 3.8557,
+      "step": 2464768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.635797644465963e-06,
+      "loss": 3.8534,
+      "step": 2465280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.627411696955443e-06,
+      "loss": 3.8522,
+      "step": 2465792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.619025749444922e-06,
+      "loss": 3.8536,
+      "step": 2466304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.610639801934403e-06,
+      "loss": 3.8257,
+      "step": 2466816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.602270233227614e-06,
+      "loss": 3.8474,
+      "step": 2467328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.593884285717094e-06,
+      "loss": 3.8572,
+      "step": 2467840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.585498338206574e-06,
+      "loss": 3.8489,
+      "step": 2468352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.577112390696054e-06,
+      "loss": 3.8495,
+      "step": 2468864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.568726443185534e-06,
+      "loss": 3.8381,
+      "step": 2469376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.560340495675014e-06,
+      "loss": 3.8437,
+      "step": 2469888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.551954548164494e-06,
+      "loss": 3.8254,
+      "step": 2470400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.543568600653974e-06,
+      "loss": 3.8584,
+      "step": 2470912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.535199031947185e-06,
+      "loss": 3.827,
+      "step": 2471424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.526813084436665e-06,
+      "loss": 3.8529,
+      "step": 2471936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.518427136926145e-06,
+      "loss": 3.8489,
+      "step": 2472448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.510041189415624e-06,
+      "loss": 3.836,
+      "step": 2472960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.501671620708835e-06,
+      "loss": 3.8426,
+      "step": 2473472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.493285673198315e-06,
+      "loss": 3.8413,
+      "step": 2473984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.484899725687795e-06,
+      "loss": 3.839,
+      "step": 2474496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.476513778177277e-06,
+      "loss": 3.8342,
+      "step": 2475008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.468127830666755e-06,
+      "loss": 3.8517,
+      "step": 2475520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.459741883156235e-06,
+      "loss": 3.8411,
+      "step": 2476032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.451355935645715e-06,
+      "loss": 3.8371,
+      "step": 2476544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.442986366938926e-06,
+      "loss": 3.8373,
+      "step": 2477056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.434600419428406e-06,
+      "loss": 3.8386,
+      "step": 2477568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.426214471917887e-06,
+      "loss": 3.8536,
+      "step": 2478080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.417828524407366e-06,
+      "loss": 3.8492,
+      "step": 2478592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.409442576896846e-06,
+      "loss": 3.8386,
+      "step": 2479104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.401073008190058e-06,
+      "loss": 3.8501,
+      "step": 2479616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.392687060679538e-06,
+      "loss": 3.8583,
+      "step": 2480128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.384301113169016e-06,
+      "loss": 3.834,
+      "step": 2480640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.375915165658496e-06,
+      "loss": 3.8503,
+      "step": 2481152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.367545596951709e-06,
+      "loss": 3.8447,
+      "step": 2481664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.359159649441189e-06,
+      "loss": 3.8432,
+      "step": 2482176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.350773701930669e-06,
+      "loss": 3.8394,
+      "step": 2482688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.342387754420149e-06,
+      "loss": 3.8448,
+      "step": 2483200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.334001806909629e-06,
+      "loss": 3.8542,
+      "step": 2483712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.325615859399107e-06,
+      "loss": 3.8606,
+      "step": 2484224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.31724629069232e-06,
+      "loss": 3.8389,
+      "step": 2484736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.3088603431818e-06,
+      "loss": 3.8212,
+      "step": 2485248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.30047439567128e-06,
+      "loss": 3.8413,
+      "step": 2485760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.29208844816076e-06,
+      "loss": 3.8452,
+      "step": 2486272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.283702500650239e-06,
+      "loss": 3.8481,
+      "step": 2486784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.275316553139719e-06,
+      "loss": 3.8414,
+      "step": 2487296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.266930605629199e-06,
+      "loss": 3.846,
+      "step": 2487808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.25856103692241e-06,
+      "loss": 3.8456,
+      "step": 2488320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.250191468215622e-06,
+      "loss": 3.8466,
+      "step": 2488832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.241805520705102e-06,
+      "loss": 3.8355,
+      "step": 2489344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.23341957319458e-06,
+      "loss": 3.8495,
+      "step": 2489856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.22503362568406e-06,
+      "loss": 3.8401,
+      "step": 2490368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.216647678173542e-06,
+      "loss": 3.8544,
+      "step": 2490880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.20826173066302e-06,
+      "loss": 3.8467,
+      "step": 2491392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.1998757831525e-06,
+      "loss": 3.8297,
+      "step": 2491904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.19148983564198e-06,
+      "loss": 3.8459,
+      "step": 2492416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.183103888131462e-06,
+      "loss": 3.848,
+      "step": 2492928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.174734319424671e-06,
+      "loss": 3.8543,
+      "step": 2493440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.166348371914151e-06,
+      "loss": 3.8429,
+      "step": 2493952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.157962424403633e-06,
+      "loss": 3.8454,
+      "step": 2494464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.149576476893111e-06,
+      "loss": 3.8344,
+      "step": 2494976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.141206908186323e-06,
+      "loss": 3.8476,
+      "step": 2495488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.132820960675803e-06,
+      "loss": 3.8308,
+      "step": 2496000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.124435013165283e-06,
+      "loss": 3.8413,
+      "step": 2496512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.116049065654762e-06,
+      "loss": 3.8457,
+      "step": 2497024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.107663118144243e-06,
+      "loss": 3.8442,
+      "step": 2497536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.099293549437454e-06,
+      "loss": 3.8335,
+      "step": 2498048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.090907601926934e-06,
+      "loss": 3.8437,
+      "step": 2498560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.082521654416414e-06,
+      "loss": 3.8379,
+      "step": 2499072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.074135706905894e-06,
+      "loss": 3.8479,
+      "step": 2499584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.065749759395374e-06,
+      "loss": 3.8423,
+      "step": 2500096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.057363811884854e-06,
+      "loss": 3.8424,
+      "step": 2500608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.048994243178065e-06,
+      "loss": 3.8546,
+      "step": 2501120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.040608295667545e-06,
+      "loss": 3.8546,
+      "step": 2501632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.032222348157025e-06,
+      "loss": 3.8466,
+      "step": 2502144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.023836400646504e-06,
+      "loss": 3.8389,
+      "step": 2502656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.015466831939715e-06,
+      "loss": 3.8449,
+      "step": 2503168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.007080884429197e-06,
+      "loss": 3.8429,
+      "step": 2503680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.998694936918675e-06,
+      "loss": 3.8399,
+      "step": 2504192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.990308989408155e-06,
+      "loss": 3.8551,
+      "step": 2504704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.981923041897635e-06,
+      "loss": 3.839,
+      "step": 2505216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.973537094387117e-06,
+      "loss": 3.8491,
+      "step": 2505728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.965151146876595e-06,
+      "loss": 3.8345,
+      "step": 2506240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.956781578169807e-06,
+      "loss": 3.8394,
+      "step": 2506752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.948395630659287e-06,
+      "loss": 3.8452,
+      "step": 2507264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.940009683148766e-06,
+      "loss": 3.8416,
+      "step": 2507776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.931623735638246e-06,
+      "loss": 3.8344,
+      "step": 2508288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.923237788127727e-06,
+      "loss": 3.863,
+      "step": 2508800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.914868219420938e-06,
+      "loss": 3.8489,
+      "step": 2509312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.906482271910416e-06,
+      "loss": 3.8427,
+      "step": 2509824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.898096324399898e-06,
+      "loss": 3.8427,
+      "step": 2510336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.889710376889378e-06,
+      "loss": 3.8469,
+      "step": 2510848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.881324429378856e-06,
+      "loss": 3.85,
+      "step": 2511360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.872938481868336e-06,
+      "loss": 3.8458,
+      "step": 2511872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.864568913161549e-06,
+      "loss": 3.8435,
+      "step": 2512384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.856182965651029e-06,
+      "loss": 3.8556,
+      "step": 2512896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.84781339694424e-06,
+      "loss": 3.8529,
+      "step": 2513408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.83942744943372e-06,
+      "loss": 3.8496,
+      "step": 2513920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.8310415019232e-06,
+      "loss": 3.8458,
+      "step": 2514432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.82265555441268e-06,
+      "loss": 3.8488,
+      "step": 2514944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.81426960690216e-06,
+      "loss": 3.8489,
+      "step": 2515456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.80590003819537e-06,
+      "loss": 3.8482,
+      "step": 2515968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.797514090684852e-06,
+      "loss": 3.8462,
+      "step": 2516480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.78912814317433e-06,
+      "loss": 3.8458,
+      "step": 2516992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.78074219566381e-06,
+      "loss": 3.8531,
+      "step": 2517504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.772372626957022e-06,
+      "loss": 3.8378,
+      "step": 2518016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.763986679446502e-06,
+      "loss": 3.8413,
+      "step": 2518528
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.979726791381836,
+      "eval_runtime": 285.3691,
+      "eval_samples_per_second": 1337.184,
+      "eval_steps_per_second": 41.788,
+      "step": 2518560
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.0406739636814202e+18,
+  "trial_name": null,
+  "trial_params": null
+}