diff --git "a/checkpoint-839509/trainer_state.json" "b/checkpoint-839509/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-839509/trainer_state.json"
@@ -0,0 +1,9947 @@
+{
+  "best_metric": 4.067577362060547,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl/lstm/1/checkpoints/checkpoint-839509",
+  "epoch": 1.0250002784396635,
+  "eval_steps": 10,
+  "global_step": 839509,
+  "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": 10.8173,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5621,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0552,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9778,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9512,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8598,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7142,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6064,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5223,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4391,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.405,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.347,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2854,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2171,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1624,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.1169,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0706,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 6.0137,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9735,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.9362,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8977,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8699,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8394,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.807,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7783,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7604,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7276,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.7055,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6716,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.667,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6342,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6179,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.6075,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.589,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714926921053515e-05,
+      "loss": 5.5594,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706540973542995e-05,
+      "loss": 5.5456,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5373,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.5153,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9681399509815164e-05,
+      "loss": 5.5032,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9673013562304644e-05,
+      "loss": 5.4829,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4577,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4464,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4322,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4251,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.4132,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.4073,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3832,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3828,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3878,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3593,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3489,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3364,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3456,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.3071,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 5.3256,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 5.2948,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 5.2928,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 5.2861,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 5.2759,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 5.2628,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.2518,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2376,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2442,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 5.2314,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 5.2213,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.2138,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.2024,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1858,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.203,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1795,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1736,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1824,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1575,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1701,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1621,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1505,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1492,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1231,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.1214,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.1192,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1184,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.1077,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0946,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0808,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0878,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0868,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0809,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0794,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0745,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0617,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0432,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0402,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0459,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0425,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0189,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0237,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.0224,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 5.0212,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 5.0117,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 5.0127,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 5.0045,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9936,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9952,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9946,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9785,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9824,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9603,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9551,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9542,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.9587,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.9513,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906945464479944e-05,
+      "loss": 4.9289,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906106869728892e-05,
+      "loss": 4.9339,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9377,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9436,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9289,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9288,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9254,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.9262,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.9102,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9102,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.9012,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.905,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8878,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8999,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8903,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8972,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8688,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8801,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8895,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8614,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.882,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8646,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.868,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8582,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8538,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.8555,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8526,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8526,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8449,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8474,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.8465,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.8267,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8271,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.818,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.8204,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.8187,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.8119,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8099,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.8102,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.7706427574157715,
+      "eval_runtime": 545.2364,
+      "eval_samples_per_second": 699.863,
+      "eval_steps_per_second": 21.871,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.819,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.8106,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.7982,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871737587978745e-05,
+      "loss": 4.8001,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870898993227693e-05,
+      "loss": 4.7934,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.87006039847664e-05,
+      "loss": 4.7726,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869221803725588e-05,
+      "loss": 4.785,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.868383208974536e-05,
+      "loss": 4.7882,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867544614223484e-05,
+      "loss": 4.7692,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866706019472432e-05,
+      "loss": 4.785,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86586742472138e-05,
+      "loss": 4.7886,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865028829970329e-05,
+      "loss": 4.7729,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864190235219277e-05,
+      "loss": 4.7658,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863351640468225e-05,
+      "loss": 4.7677,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862513045717173e-05,
+      "loss": 4.767,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861674450966121e-05,
+      "loss": 4.7616,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860835856215069e-05,
+      "loss": 4.7565,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.859997261464017e-05,
+      "loss": 4.756,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.859158666712964e-05,
+      "loss": 4.7509,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8583233477226585e-05,
+      "loss": 4.7549,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574847529716065e-05,
+      "loss": 4.7551,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566461582205545e-05,
+      "loss": 4.7543,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558075634695025e-05,
+      "loss": 4.7496,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8549689687184505e-05,
+      "loss": 4.7474,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854132011847772e-05,
+      "loss": 4.7419,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85329341709672e-05,
+      "loss": 4.7387,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.852454822345668e-05,
+      "loss": 4.7425,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.851616227594616e-05,
+      "loss": 4.723,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850777632843564e-05,
+      "loss": 4.7354,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8499390380925114e-05,
+      "loss": 4.7177,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8491004433414594e-05,
+      "loss": 4.722,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8482618485904074e-05,
+      "loss": 4.7319,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847424891719729e-05,
+      "loss": 4.7296,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.846586296968676e-05,
+      "loss": 4.715,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845747702217624e-05,
+      "loss": 4.707,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844910745346946e-05,
+      "loss": 4.7239,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844072150595894e-05,
+      "loss": 4.7108,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843233555844842e-05,
+      "loss": 4.7093,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84239496109379e-05,
+      "loss": 4.7055,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841556366342738e-05,
+      "loss": 4.6932,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840717771591686e-05,
+      "loss": 4.6894,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839879176840634e-05,
+      "loss": 4.6913,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839040582089582e-05,
+      "loss": 4.6954,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83820198733853e-05,
+      "loss": 4.6896,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837363392587478e-05,
+      "loss": 4.6924,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836524797836426e-05,
+      "loss": 4.6844,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835686203085374e-05,
+      "loss": 4.6879,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834849246214695e-05,
+      "loss": 4.6993,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834010651463643e-05,
+      "loss": 4.6794,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833172056712591e-05,
+      "loss": 4.6801,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832333461961539e-05,
+      "loss": 4.6698,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8314948672104874e-05,
+      "loss": 4.6902,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830657910339808e-05,
+      "loss": 4.6616,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829819315588756e-05,
+      "loss": 4.6838,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828982358718077e-05,
+      "loss": 4.6604,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828143763967025e-05,
+      "loss": 4.6699,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.827305169215973e-05,
+      "loss": 4.6682,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.826466574464921e-05,
+      "loss": 4.6641,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.825627979713869e-05,
+      "loss": 4.6559,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.824789384962817e-05,
+      "loss": 4.6559,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823950790211765e-05,
+      "loss": 4.6452,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231121954607125e-05,
+      "loss": 4.6602,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.822273600709661e-05,
+      "loss": 4.6511,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.821435005958609e-05,
+      "loss": 4.6454,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.82059804908793e-05,
+      "loss": 4.6438,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.819759454336878e-05,
+      "loss": 4.6443,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818920859585826e-05,
+      "loss": 4.6261,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818082264834774e-05,
+      "loss": 4.65,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817245307964095e-05,
+      "loss": 4.6354,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816406713213043e-05,
+      "loss": 4.6349,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.815568118461991e-05,
+      "loss": 4.6435,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814729523710939e-05,
+      "loss": 4.6292,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813890928959887e-05,
+      "loss": 4.6431,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813053972089208e-05,
+      "loss": 4.6458,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8122153773381565e-05,
+      "loss": 4.6363,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8113767825871045e-05,
+      "loss": 4.6383,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8105381878360525e-05,
+      "loss": 4.6205,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8096995930850005e-05,
+      "loss": 4.6156,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8088609983339485e-05,
+      "loss": 4.6269,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8080240414632694e-05,
+      "loss": 4.6279,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8071854467122174e-05,
+      "loss": 4.6204,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8063468519611654e-05,
+      "loss": 4.6159,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8055082572101134e-05,
+      "loss": 4.5988,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8046696624590614e-05,
+      "loss": 4.61,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8038310677080094e-05,
+      "loss": 4.6138,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8029924729569574e-05,
+      "loss": 4.6127,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8021538782059054e-05,
+      "loss": 4.6194,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8013152834548534e-05,
+      "loss": 4.6174,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800478326584174e-05,
+      "loss": 4.607,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799639731833123e-05,
+      "loss": 4.5994,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798802774962444e-05,
+      "loss": 4.5926,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797964180211392e-05,
+      "loss": 4.599,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.79712558546034e-05,
+      "loss": 4.6017,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796286990709288e-05,
+      "loss": 4.5893,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795448395958236e-05,
+      "loss": 4.5903,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794611439087557e-05,
+      "loss": 4.5981,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793772844336505e-05,
+      "loss": 4.6013,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792934249585453e-05,
+      "loss": 4.5899,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792095654834401e-05,
+      "loss": 4.596,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791257060083349e-05,
+      "loss": 4.5978,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.790418465332297e-05,
+      "loss": 4.5883,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789579870581245e-05,
+      "loss": 4.585,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788741275830193e-05,
+      "loss": 4.5951,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7879043189595136e-05,
+      "loss": 4.5802,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7870657242084616e-05,
+      "loss": 4.5931,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.786228767337783e-05,
+      "loss": 4.573,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.785390172586731e-05,
+      "loss": 4.5631,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845515778356785e-05,
+      "loss": 4.5704,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837129830846265e-05,
+      "loss": 4.5787,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782876026213948e-05,
+      "loss": 4.572,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782037431462896e-05,
+      "loss": 4.5589,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7811988367118434e-05,
+      "loss": 4.559,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780360241960792e-05,
+      "loss": 4.5672,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77952164720974e-05,
+      "loss": 4.5783,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778683052458688e-05,
+      "loss": 4.565,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777844457707636e-05,
+      "loss": 4.5736,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777005862956584e-05,
+      "loss": 4.5694,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776168906085905e-05,
+      "loss": 4.5658,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775331949215226e-05,
+      "loss": 4.5613,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774493354464174e-05,
+      "loss": 4.5578,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773654759713122e-05,
+      "loss": 4.5541,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77281616496207e-05,
+      "loss": 4.5645,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771977570211018e-05,
+      "loss": 4.5491,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771138975459966e-05,
+      "loss": 4.5512,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.770300380708914e-05,
+      "loss": 4.5553,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769461785957862e-05,
+      "loss": 4.5625,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7686248290871835e-05,
+      "loss": 4.5395,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7677862343361315e-05,
+      "loss": 4.543,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7669476395850795e-05,
+      "loss": 4.5655,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7661090448340274e-05,
+      "loss": 4.5387,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7652720879633484e-05,
+      "loss": 4.5605,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764435131092669e-05,
+      "loss": 4.5427,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.763596536341617e-05,
+      "loss": 4.5539,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762757941590565e-05,
+      "loss": 4.5414,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761919346839513e-05,
+      "loss": 4.537,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761080752088461e-05,
+      "loss": 4.5451,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760242157337409e-05,
+      "loss": 4.5409,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759405200466731e-05,
+      "loss": 4.5498,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758566605715679e-05,
+      "loss": 4.5441,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757728010964627e-05,
+      "loss": 4.5433,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756889416213575e-05,
+      "loss": 4.5408,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756052459342896e-05,
+      "loss": 4.5256,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755213864591844e-05,
+      "loss": 4.536,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754375269840792e-05,
+      "loss": 4.5226,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75353667508974e-05,
+      "loss": 4.5275,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7526997182190606e-05,
+      "loss": 4.5307,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7518611234680086e-05,
+      "loss": 4.518,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7510225287169566e-05,
+      "loss": 4.5247,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7501839339659046e-05,
+      "loss": 4.5218,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.488372802734375,
+      "eval_runtime": 546.2477,
+      "eval_samples_per_second": 698.568,
+      "eval_steps_per_second": 21.831,
+      "step": 152638
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7493453392148526e-05,
+      "loss": 4.5335,
+      "step": 153088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.748508382344174e-05,
+      "loss": 4.5308,
+      "step": 153600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.747669787593122e-05,
+      "loss": 4.5211,
+      "step": 154112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.5176,
+      "step": 154624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.5188,
+      "step": 155136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745155641220339e-05,
+      "loss": 4.5028,
+      "step": 155648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.744317046469287e-05,
+      "loss": 4.5165,
+      "step": 156160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.743480089598608e-05,
+      "loss": 4.5148,
+      "step": 156672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.742641494847556e-05,
+      "loss": 4.4999,
+      "step": 157184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.741802900096504e-05,
+      "loss": 4.5196,
+      "step": 157696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.740964305345452e-05,
+      "loss": 4.5243,
+      "step": 158208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7401257105944e-05,
+      "loss": 4.5079,
+      "step": 158720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.739287115843348e-05,
+      "loss": 4.5009,
+      "step": 159232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7384485210922966e-05,
+      "loss": 4.5046,
+      "step": 159744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7376099263412446e-05,
+      "loss": 4.508,
+      "step": 160256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7367729694705656e-05,
+      "loss": 4.5042,
+      "step": 160768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7359343747195135e-05,
+      "loss": 4.4989,
+      "step": 161280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7350957799684615e-05,
+      "loss": 4.4989,
+      "step": 161792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4999,
+      "step": 162304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.5027,
+      "step": 162816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7325816335956784e-05,
+      "loss": 4.5034,
+      "step": 163328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7317430388446264e-05,
+      "loss": 4.5055,
+      "step": 163840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730904444093574e-05,
+      "loss": 4.5033,
+      "step": 164352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730065849342522e-05,
+      "loss": 4.4952,
+      "step": 164864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7292288924718433e-05,
+      "loss": 4.5005,
+      "step": 165376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.728390297720791e-05,
+      "loss": 4.4946,
+      "step": 165888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.727551702969739e-05,
+      "loss": 4.4982,
+      "step": 166400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.726713108218687e-05,
+      "loss": 4.4837,
+      "step": 166912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.725876151348009e-05,
+      "loss": 4.4946,
+      "step": 167424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.725037556596957e-05,
+      "loss": 4.4793,
+      "step": 167936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.724198961845904e-05,
+      "loss": 4.4858,
+      "step": 168448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.723360367094852e-05,
+      "loss": 4.4955,
+      "step": 168960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7225217723438e-05,
+      "loss": 4.4939,
+      "step": 169472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.721684815473121e-05,
+      "loss": 4.4818,
+      "step": 169984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.476,
+      "step": 170496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.498,
+      "step": 171008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4778,
+      "step": 171520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.718330436468914e-05,
+      "loss": 4.4861,
+      "step": 172032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.717491841717862e-05,
+      "loss": 4.4812,
+      "step": 172544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.71665324696681e-05,
+      "loss": 4.4669,
+      "step": 173056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.715814652215758e-05,
+      "loss": 4.4645,
+      "step": 173568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714977695345079e-05,
+      "loss": 4.4655,
+      "step": 174080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714139100594027e-05,
+      "loss": 4.4806,
+      "step": 174592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7133021437233476e-05,
+      "loss": 4.4716,
+      "step": 175104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7124635489722956e-05,
+      "loss": 4.4708,
+      "step": 175616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7116249542212436e-05,
+      "loss": 4.4651,
+      "step": 176128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7107863594701916e-05,
+      "loss": 4.4686,
+      "step": 176640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7099477647191396e-05,
+      "loss": 4.4834,
+      "step": 177152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7091091699680876e-05,
+      "loss": 4.4632,
+      "step": 177664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.708272213097409e-05,
+      "loss": 4.4633,
+      "step": 178176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.707433618346357e-05,
+      "loss": 4.459,
+      "step": 178688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.706595023595305e-05,
+      "loss": 4.4801,
+      "step": 179200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.705756428844253e-05,
+      "loss": 4.4513,
+      "step": 179712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704917834093201e-05,
+      "loss": 4.4739,
+      "step": 180224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704079239342149e-05,
+      "loss": 4.4501,
+      "step": 180736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.70324228247147e-05,
+      "loss": 4.4635,
+      "step": 181248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.702403687720418e-05,
+      "loss": 4.4644,
+      "step": 181760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.701565092969366e-05,
+      "loss": 4.453,
+      "step": 182272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.700726498218314e-05,
+      "loss": 4.4572,
+      "step": 182784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.699887903467262e-05,
+      "loss": 4.4511,
+      "step": 183296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.69904930871621e-05,
+      "loss": 4.4369,
+      "step": 183808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6982107139651573e-05,
+      "loss": 4.4557,
+      "step": 184320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.697373757094479e-05,
+      "loss": 4.4539,
+      "step": 184832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6965351623434276e-05,
+      "loss": 4.4444,
+      "step": 185344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6956965675923756e-05,
+      "loss": 4.4472,
+      "step": 185856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694857972841323e-05,
+      "loss": 4.4476,
+      "step": 186368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694019378090271e-05,
+      "loss": 4.429,
+      "step": 186880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.693180783339219e-05,
+      "loss": 4.4563,
+      "step": 187392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.437,
+      "step": 187904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.441,
+      "step": 188416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.690666636966436e-05,
+      "loss": 4.4512,
+      "step": 188928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.689828042215384e-05,
+      "loss": 4.4379,
+      "step": 189440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.688989447464332e-05,
+      "loss": 4.4489,
+      "step": 189952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4582,
+      "step": 190464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4445,
+      "step": 190976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4521,
+      "step": 191488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6856367063404974e-05,
+      "loss": 4.4343,
+      "step": 192000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6847981115894454e-05,
+      "loss": 4.4262,
+      "step": 192512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6839595168383934e-05,
+      "loss": 4.4387,
+      "step": 193024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6831209220873414e-05,
+      "loss": 4.4461,
+      "step": 193536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6822823273362894e-05,
+      "loss": 4.437,
+      "step": 194048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6814437325852373e-05,
+      "loss": 4.4288,
+      "step": 194560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6806051378341853e-05,
+      "loss": 4.4202,
+      "step": 195072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.4244,
+      "step": 195584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678929586212454e-05,
+      "loss": 4.4274,
+      "step": 196096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.431,
+      "step": 196608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.67725239671035e-05,
+      "loss": 4.4403,
+      "step": 197120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.676415439839671e-05,
+      "loss": 4.4375,
+      "step": 197632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.675576845088619e-05,
+      "loss": 4.4276,
+      "step": 198144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.423,
+      "step": 198656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.4165,
+      "step": 199168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4228,
+      "step": 199680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.672222466084412e-05,
+      "loss": 4.4209,
+      "step": 200192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.671383871333359e-05,
+      "loss": 4.4166,
+      "step": 200704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.670545276582307e-05,
+      "loss": 4.4163,
+      "step": 201216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.669706681831255e-05,
+      "loss": 4.426,
+      "step": 201728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668868087080203e-05,
+      "loss": 4.424,
+      "step": 202240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668031130209524e-05,
+      "loss": 4.4243,
+      "step": 202752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.667192535458472e-05,
+      "loss": 4.4239,
+      "step": 203264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.66635394070742e-05,
+      "loss": 4.4267,
+      "step": 203776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.665515345956368e-05,
+      "loss": 4.4198,
+      "step": 204288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6646783890856896e-05,
+      "loss": 4.4136,
+      "step": 204800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.4292,
+      "step": 205312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.663002837463959e-05,
+      "loss": 4.4085,
+      "step": 205824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.4311,
+      "step": 206336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.4025,
+      "step": 206848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.3963,
+      "step": 207360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.4052,
+      "step": 207872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6588098637086985e-05,
+      "loss": 4.4194,
+      "step": 208384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6579712689576465e-05,
+      "loss": 4.4091,
+      "step": 208896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6571326742065945e-05,
+      "loss": 4.3951,
+      "step": 209408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6562940794555425e-05,
+      "loss": 4.3962,
+      "step": 209920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6554554847044905e-05,
+      "loss": 4.4031,
+      "step": 210432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6546168899534385e-05,
+      "loss": 4.418,
+      "step": 210944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6537782952023865e-05,
+      "loss": 4.4033,
+      "step": 211456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.4126,
+      "step": 211968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652104381461029e-05,
+      "loss": 4.4087,
+      "step": 212480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.651265786709977e-05,
+      "loss": 4.406,
+      "step": 212992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.650427191958925e-05,
+      "loss": 4.4093,
+      "step": 213504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.649588597207873e-05,
+      "loss": 4.3976,
+      "step": 214016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.648750002456821e-05,
+      "loss": 4.3992,
+      "step": 214528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647911407705769e-05,
+      "loss": 4.4079,
+      "step": 215040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647072812954717e-05,
+      "loss": 4.3909,
+      "step": 215552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.646234218203665e-05,
+      "loss": 4.3972,
+      "step": 216064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.645397261332986e-05,
+      "loss": 4.401,
+      "step": 216576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.4045,
+      "step": 217088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.3891,
+      "step": 217600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6428831149602034e-05,
+      "loss": 4.3887,
+      "step": 218112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6420445202091514e-05,
+      "loss": 4.4135,
+      "step": 218624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6412059254580994e-05,
+      "loss": 4.3912,
+      "step": 219136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6403673307070474e-05,
+      "loss": 4.4041,
+      "step": 219648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.639528735955995e-05,
+      "loss": 4.3949,
+      "step": 220160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.638690141204943e-05,
+      "loss": 4.4027,
+      "step": 220672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637851546453891e-05,
+      "loss": 4.3912,
+      "step": 221184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637012951702839e-05,
+      "loss": 4.3886,
+      "step": 221696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.3994,
+      "step": 222208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6353374000811076e-05,
+      "loss": 4.3889,
+      "step": 222720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6344988053300556e-05,
+      "loss": 4.4018,
+      "step": 223232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.633661848459377e-05,
+      "loss": 4.399,
+      "step": 223744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.632823253708325e-05,
+      "loss": 4.3967,
+      "step": 224256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.631984658957273e-05,
+      "loss": 4.3914,
+      "step": 224768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.631146064206221e-05,
+      "loss": 4.3834,
+      "step": 225280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.630307469455169e-05,
+      "loss": 4.3895,
+      "step": 225792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.629468874704117e-05,
+      "loss": 4.3791,
+      "step": 226304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.628630279953065e-05,
+      "loss": 4.3838,
+      "step": 226816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.627791685202013e-05,
+      "loss": 4.3844,
+      "step": 227328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626953090450961e-05,
+      "loss": 4.3776,
+      "step": 227840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626114495699909e-05,
+      "loss": 4.3796,
+      "step": 228352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.62527753882923e-05,
+      "loss": 4.3799,
+      "step": 228864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.354189872741699,
+      "eval_runtime": 546.5547,
+      "eval_samples_per_second": 698.175,
+      "eval_steps_per_second": 21.818,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624438944078178e-05,
+      "loss": 4.3894,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623600349327126e-05,
+      "loss": 4.3938,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622761754576074e-05,
+      "loss": 4.3809,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621923159825022e-05,
+      "loss": 4.3769,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.62108456507397e-05,
+      "loss": 4.3806,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620245970322919e-05,
+      "loss": 4.362,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619407375571867e-05,
+      "loss": 4.3777,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618568780820814e-05,
+      "loss": 4.379,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617730186069762e-05,
+      "loss": 4.3667,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6168932291990836e-05,
+      "loss": 4.3811,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616054634448031e-05,
+      "loss": 4.3879,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615216039696979e-05,
+      "loss": 4.3686,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614377444945927e-05,
+      "loss": 4.3683,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613538850194875e-05,
+      "loss": 4.3707,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612700255443823e-05,
+      "loss": 4.3748,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611861660692771e-05,
+      "loss": 4.3688,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611023065941719e-05,
+      "loss": 4.3635,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610184471190667e-05,
+      "loss": 4.3654,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093475143199885e-05,
+      "loss": 4.3667,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085089195689365e-05,
+      "loss": 4.3738,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6076703248178845e-05,
+      "loss": 4.3706,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6068317300668325e-05,
+      "loss": 4.3753,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6059931353157805e-05,
+      "loss": 4.3721,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6051545405647285e-05,
+      "loss": 4.3666,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6043159458136764e-05,
+      "loss": 4.3652,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6034773510626244e-05,
+      "loss": 4.3653,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6026387563115724e-05,
+      "loss": 4.3701,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60180016156052e-05,
+      "loss": 4.3575,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6009632046898413e-05,
+      "loss": 4.3661,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6001246099387893e-05,
+      "loss": 4.3523,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599286015187737e-05,
+      "loss": 4.3565,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598449058317059e-05,
+      "loss": 4.3703,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597610463566007e-05,
+      "loss": 4.3659,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596771868814955e-05,
+      "loss": 4.3581,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595933274063903e-05,
+      "loss": 4.352,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.59509467931285e-05,
+      "loss": 4.3663,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594256084561798e-05,
+      "loss": 4.3538,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593417489810746e-05,
+      "loss": 4.3646,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.592578895059694e-05,
+      "loss": 4.3548,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591740300308642e-05,
+      "loss": 4.3442,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.59090170555759e-05,
+      "loss": 4.3437,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590064748686911e-05,
+      "loss": 4.3455,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589226153935859e-05,
+      "loss": 4.3555,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588387559184808e-05,
+      "loss": 4.3468,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587548964433756e-05,
+      "loss": 4.3488,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586712007563077e-05,
+      "loss": 4.3452,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585873412812025e-05,
+      "loss": 4.3494,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585034818060973e-05,
+      "loss": 4.3605,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584196223309921e-05,
+      "loss": 4.342,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583357628558869e-05,
+      "loss": 4.3423,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582519033807817e-05,
+      "loss": 4.3463,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581680439056765e-05,
+      "loss": 4.3567,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5808418443057127e-05,
+      "loss": 4.3344,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5800048874350336e-05,
+      "loss": 4.3516,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5791662926839816e-05,
+      "loss": 4.3329,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783276979329296e-05,
+      "loss": 4.3415,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5774891031818776e-05,
+      "loss": 4.3468,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576653784191572e-05,
+      "loss": 4.3334,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57581518944052e-05,
+      "loss": 4.347,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574976594689468e-05,
+      "loss": 4.3389,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574137999938416e-05,
+      "loss": 4.3219,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.573299405187364e-05,
+      "loss": 4.3363,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.572462448316685e-05,
+      "loss": 4.337,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.571623853565633e-05,
+      "loss": 4.3356,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.570785258814581e-05,
+      "loss": 4.3327,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569946664063529e-05,
+      "loss": 4.3332,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569108069312477e-05,
+      "loss": 4.3157,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568269474561425e-05,
+      "loss": 4.3427,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674325176907465e-05,
+      "loss": 4.3236,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5665939229396945e-05,
+      "loss": 4.3279,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5657569660690154e-05,
+      "loss": 4.3391,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5649183713179634e-05,
+      "loss": 4.3266,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5640797765669114e-05,
+      "loss": 4.3361,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5632411818158594e-05,
+      "loss": 4.344,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5624025870648074e-05,
+      "loss": 4.3349,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561565630194128e-05,
+      "loss": 4.3427,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560727035443076e-05,
+      "loss": 4.3235,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559888440692024e-05,
+      "loss": 4.3186,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559049845940972e-05,
+      "loss": 4.3268,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55821125118992e-05,
+      "loss": 4.3385,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557372656438868e-05,
+      "loss": 4.3271,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556534061687816e-05,
+      "loss": 4.3198,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555695466936764e-05,
+      "loss": 4.3105,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554858510066086e-05,
+      "loss": 4.3204,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554019915315033e-05,
+      "loss": 4.3183,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553181320563981e-05,
+      "loss": 4.3221,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552342725812929e-05,
+      "loss": 4.3288,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551504131061877e-05,
+      "loss": 4.334,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550665536310825e-05,
+      "loss": 4.3216,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549826941559773e-05,
+      "loss": 4.3195,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548989984689094e-05,
+      "loss": 4.309,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548151389938042e-05,
+      "loss": 4.3138,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54731279518699e-05,
+      "loss": 4.3189,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546474200435939e-05,
+      "loss": 4.3126,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545635605684887e-05,
+      "loss": 4.3085,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544797010933835e-05,
+      "loss": 4.3228,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543958416182783e-05,
+      "loss": 4.3159,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543119821431731e-05,
+      "loss": 4.3262,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5422828645610516e-05,
+      "loss": 4.3164,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5414442698099996e-05,
+      "loss": 4.3242,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5406073129393205e-05,
+      "loss": 4.3154,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5397687181882685e-05,
+      "loss": 4.31,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389301234372165e-05,
+      "loss": 4.328,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5380915286861645e-05,
+      "loss": 4.3068,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372529339351125e-05,
+      "loss": 4.3303,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5364143391840605e-05,
+      "loss": 4.3016,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5355757444330085e-05,
+      "loss": 4.2976,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534737149681957e-05,
+      "loss": 4.3033,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533900192811278e-05,
+      "loss": 4.3188,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533063235940599e-05,
+      "loss": 4.3097,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532224641189547e-05,
+      "loss": 4.2946,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531386046438495e-05,
+      "loss": 4.2941,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.530547451687443e-05,
+      "loss": 4.3027,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529710494816764e-05,
+      "loss": 4.3206,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528871900065712e-05,
+      "loss": 4.3046,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52803330531466e-05,
+      "loss": 4.3176,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527194710563608e-05,
+      "loss": 4.3086,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526356115812556e-05,
+      "loss": 4.3094,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.525517521061504e-05,
+      "loss": 4.3105,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524678926310452e-05,
+      "loss": 4.3001,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238403315594e-05,
+      "loss": 4.3013,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5230033746887215e-05,
+      "loss": 4.3081,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5221647799376694e-05,
+      "loss": 4.2952,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5213278230669904e-05,
+      "loss": 4.3016,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5204892283159384e-05,
+      "loss": 4.3066,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519652271445259e-05,
+      "loss": 4.3043,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518813676694207e-05,
+      "loss": 4.2946,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517975081943155e-05,
+      "loss": 4.2942,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517136487192103e-05,
+      "loss": 4.3154,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516297892441051e-05,
+      "loss": 4.2976,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515459297689999e-05,
+      "loss": 4.3084,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514620702938947e-05,
+      "loss": 4.3015,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513782108187895e-05,
+      "loss": 4.3068,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512943513436843e-05,
+      "loss": 4.2982,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512104918685791e-05,
+      "loss": 4.2981,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.511266323934739e-05,
+      "loss": 4.3029,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510427729183687e-05,
+      "loss": 4.2965,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509590772313008e-05,
+      "loss": 4.3116,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508752177561956e-05,
+      "loss": 4.3008,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507913582810904e-05,
+      "loss": 4.3056,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507076625940225e-05,
+      "loss": 4.302,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506238031189173e-05,
+      "loss": 4.2914,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505399436438121e-05,
+      "loss": 4.2963,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50456084168707e-05,
+      "loss": 4.2877,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503722246936018e-05,
+      "loss": 4.293,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502883652184966e-05,
+      "loss": 4.2916,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502045057433914e-05,
+      "loss": 4.291,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501206462682862e-05,
+      "loss": 4.2846,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5003695058121826e-05,
+      "loss": 4.2923,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.270674228668213,
+      "eval_runtime": 545.1308,
+      "eval_samples_per_second": 699.999,
+      "eval_steps_per_second": 21.875,
+      "step": 305276
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4995309110611306e-05,
+      "loss": 4.2996,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.3004,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.2917,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.2845,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.2946,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.2724,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.2906,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.2865,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.28,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491986834062409e-05,
+      "loss": 4.2908,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491148239311357e-05,
+      "loss": 4.2999,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490309644560305e-05,
+      "loss": 4.2791,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489471049809253e-05,
+      "loss": 4.2842,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4886324550582003e-05,
+      "loss": 4.2821,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4877938603071483e-05,
+      "loss": 4.2884,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486955265556096e-05,
+      "loss": 4.2777,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486116670805044e-05,
+      "loss": 4.2773,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485279713934365e-05,
+      "loss": 4.2781,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.484441119183313e-05,
+      "loss": 4.2789,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.483602524432262e-05,
+      "loss": 4.2904,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.48276392968121e-05,
+      "loss": 4.2828,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481925334930158e-05,
+      "loss": 4.289,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481086740179106e-05,
+      "loss": 4.2888,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.480249783308427e-05,
+      "loss": 4.2807,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.479411188557375e-05,
+      "loss": 4.2772,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.478574231686696e-05,
+      "loss": 4.2789,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.477735636935644e-05,
+      "loss": 4.2853,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476897042184592e-05,
+      "loss": 4.2729,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.47605844743354e-05,
+      "loss": 4.2781,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475219852682488e-05,
+      "loss": 4.2747,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474381257931436e-05,
+      "loss": 4.2666,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473542663180384e-05,
+      "loss": 4.2829,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472705706309705e-05,
+      "loss": 4.2869,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471867111558653e-05,
+      "loss": 4.2762,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471028516807601e-05,
+      "loss": 4.2683,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470189922056549e-05,
+      "loss": 4.2774,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.46935296518587e-05,
+      "loss": 4.2767,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468514370434818e-05,
+      "loss": 4.2769,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467675775683766e-05,
+      "loss": 4.2707,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466837180932714e-05,
+      "loss": 4.2622,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.465998586181662e-05,
+      "loss": 4.2614,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.46515999143061e-05,
+      "loss": 4.2622,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.464323034559931e-05,
+      "loss": 4.2743,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.463484439808879e-05,
+      "loss": 4.2677,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.462645845057827e-05,
+      "loss": 4.2671,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.461807250306775e-05,
+      "loss": 4.2647,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460968655555724e-05,
+      "loss": 4.2664,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460130060804672e-05,
+      "loss": 4.2827,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.459291466053619e-05,
+      "loss": 4.2628,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4584545091829406e-05,
+      "loss": 4.2615,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576159144318886e-05,
+      "loss": 4.2679,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4567773196808366e-05,
+      "loss": 4.2713,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455938724929784e-05,
+      "loss": 4.2589,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455100130178732e-05,
+      "loss": 4.2677,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45426153542768e-05,
+      "loss": 4.2589,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.262,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452585983805949e-05,
+      "loss": 4.2673,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4517473890548975e-05,
+      "loss": 4.2523,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4509087943038455e-05,
+      "loss": 4.2702,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4500701995527935e-05,
+      "loss": 4.2569,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4492316048017415e-05,
+      "loss": 4.2456,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4483930100506895e-05,
+      "loss": 4.2535,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4475544152996375e-05,
+      "loss": 4.2603,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4467158205485855e-05,
+      "loss": 4.2632,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4458788636779064e-05,
+      "loss": 4.2538,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4450402689268544e-05,
+      "loss": 4.2588,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4442016741758024e-05,
+      "loss": 4.2344,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.443364717305123e-05,
+      "loss": 4.2659,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.442526122554071e-05,
+      "loss": 4.2481,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.441687527803019e-05,
+      "loss": 4.2494,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440848933051967e-05,
+      "loss": 4.2633,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440010338300916e-05,
+      "loss": 4.2516,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.439171743549864e-05,
+      "loss": 4.2566,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.438334786679185e-05,
+      "loss": 4.2659,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.437496191928133e-05,
+      "loss": 4.2594,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.436657597177081e-05,
+      "loss": 4.2633,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435819002426029e-05,
+      "loss": 4.2492,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434980407674977e-05,
+      "loss": 4.2418,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434141812923925e-05,
+      "loss": 4.2535,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433303218172873e-05,
+      "loss": 4.261,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43246462342182e-05,
+      "loss": 4.2512,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431626028670768e-05,
+      "loss": 4.247,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43078907180009e-05,
+      "loss": 4.2344,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429950477049038e-05,
+      "loss": 4.2475,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429111882297986e-05,
+      "loss": 4.2416,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428273287546934e-05,
+      "loss": 4.2519,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.427434692795882e-05,
+      "loss": 4.2501,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.426599373805576e-05,
+      "loss": 4.2618,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.425760779054524e-05,
+      "loss": 4.2449,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424922184303472e-05,
+      "loss": 4.2483,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.42408358955242e-05,
+      "loss": 4.2342,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4232449948013675e-05,
+      "loss": 4.2417,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4224064000503155e-05,
+      "loss": 4.2453,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4215678052992635e-05,
+      "loss": 4.2382,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4207292105482115e-05,
+      "loss": 4.238,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419893891557907e-05,
+      "loss": 4.2513,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419055296806855e-05,
+      "loss": 4.2391,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.418216702055803e-05,
+      "loss": 4.2594,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.41737810730475e-05,
+      "loss": 4.2419,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4165411504340716e-05,
+      "loss": 4.2506,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4157025556830196e-05,
+      "loss": 4.2414,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4148639609319676e-05,
+      "loss": 4.2389,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414025366180915e-05,
+      "loss": 4.2556,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4131884093102365e-05,
+      "loss": 4.2361,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4123498145591845e-05,
+      "loss": 4.2576,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411511219808132e-05,
+      "loss": 4.2332,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4106726250570805e-05,
+      "loss": 4.2303,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4098340303060284e-05,
+      "loss": 4.2295,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4089954355549764e-05,
+      "loss": 4.2463,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408160116564671e-05,
+      "loss": 4.2427,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407321521813619e-05,
+      "loss": 4.2296,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406482927062567e-05,
+      "loss": 4.2225,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405644332311514e-05,
+      "loss": 4.2337,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404805737560462e-05,
+      "loss": 4.2507,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40396714280941e-05,
+      "loss": 4.2335,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403130185938732e-05,
+      "loss": 4.248,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.402291591187679e-05,
+      "loss": 4.2387,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401452996436627e-05,
+      "loss": 4.2415,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400614401685576e-05,
+      "loss": 4.2428,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.399775806934524e-05,
+      "loss": 4.2301,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398938850063845e-05,
+      "loss": 4.2313,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398100255312793e-05,
+      "loss": 4.2374,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397261660561741e-05,
+      "loss": 4.229,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396423065810689e-05,
+      "loss": 4.2329,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.395584471059637e-05,
+      "loss": 4.2372,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394745876308585e-05,
+      "loss": 4.2362,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393907281557533e-05,
+      "loss": 4.2272,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393068686806481e-05,
+      "loss": 4.2286,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.392230092055429e-05,
+      "loss": 4.2431,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.391391497304377e-05,
+      "loss": 4.2277,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390552902553325e-05,
+      "loss": 4.2414,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389714307802273e-05,
+      "loss": 4.2358,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3888773509315936e-05,
+      "loss": 4.2373,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388040394060915e-05,
+      "loss": 4.2323,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.387201799309863e-05,
+      "loss": 4.2289,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.386363204558811e-05,
+      "loss": 4.2358,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.385524609807759e-05,
+      "loss": 4.2288,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.384686015056707e-05,
+      "loss": 4.244,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383847420305655e-05,
+      "loss": 4.2376,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383008825554603e-05,
+      "loss": 4.2386,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3821702308035505e-05,
+      "loss": 4.2361,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381333273932872e-05,
+      "loss": 4.2262,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38049467918182e-05,
+      "loss": 4.2285,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379656084430768e-05,
+      "loss": 4.2224,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378817489679716e-05,
+      "loss": 4.2261,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377978894928664e-05,
+      "loss": 4.2266,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377140300177612e-05,
+      "loss": 4.2252,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.37630170542656e-05,
+      "loss": 4.2148,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.375463110675508e-05,
+      "loss": 4.2284,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.213342666625977,
+      "eval_runtime": 543.7311,
+      "eval_samples_per_second": 701.801,
+      "eval_steps_per_second": 21.932,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374626153804829e-05,
+      "loss": 4.24,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373787559053777e-05,
+      "loss": 4.2331,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372948964302725e-05,
+      "loss": 4.2285,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372110369551673e-05,
+      "loss": 4.2234,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371271774800621e-05,
+      "loss": 4.2237,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370433180049569e-05,
+      "loss": 4.2124,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369594585298517e-05,
+      "loss": 4.2258,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368755990547465e-05,
+      "loss": 4.2155,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367917395796413e-05,
+      "loss": 4.2218,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367078801045361e-05,
+      "loss": 4.2282,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366240206294309e-05,
+      "loss": 4.2312,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3654016115432576e-05,
+      "loss": 4.2147,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364563016792205e-05,
+      "loss": 4.2231,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363724422041153e-05,
+      "loss": 4.2184,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362885827290101e-05,
+      "loss": 4.2233,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3620488704194225e-05,
+      "loss": 4.215,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36121027566837e-05,
+      "loss": 4.2129,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360371680917318e-05,
+      "loss": 4.2159,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359533086166266e-05,
+      "loss": 4.22,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358694491415214e-05,
+      "loss": 4.2275,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357855896664162e-05,
+      "loss": 4.2192,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35701730191311e-05,
+      "loss": 4.2302,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356178707162058e-05,
+      "loss": 4.2226,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355341750291379e-05,
+      "loss": 4.2216,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354503155540327e-05,
+      "loss": 4.2149,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353664560789275e-05,
+      "loss": 4.2163,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352825966038223e-05,
+      "loss": 4.2256,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351989009167544e-05,
+      "loss": 4.2125,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351150414416492e-05,
+      "loss": 4.2184,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35031181966544e-05,
+      "loss": 4.2121,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349474862794761e-05,
+      "loss": 4.2043,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348636268043709e-05,
+      "loss": 4.227,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347797673292657e-05,
+      "loss": 4.2204,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346959078541605e-05,
+      "loss": 4.2202,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346120483790553e-05,
+      "loss": 4.2049,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345281889039501e-05,
+      "loss": 4.2214,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344444932168823e-05,
+      "loss": 4.2121,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343606337417771e-05,
+      "loss": 4.2191,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342767742666719e-05,
+      "loss": 4.212,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341929147915667e-05,
+      "loss": 4.202,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3410921910449876e-05,
+      "loss": 4.202,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402535962939356e-05,
+      "loss": 4.1998,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394150015428836e-05,
+      "loss": 4.2196,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385764067918316e-05,
+      "loss": 4.2074,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377378120407796e-05,
+      "loss": 4.2034,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3368992172897276e-05,
+      "loss": 4.2104,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360606225386756e-05,
+      "loss": 4.2084,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335222027787623e-05,
+      "loss": 4.2211,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3343850709169445e-05,
+      "loss": 4.201,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333546476165893e-05,
+      "loss": 4.2051,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332707881414841e-05,
+      "loss": 4.2094,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3318692866637885e-05,
+      "loss": 4.2109,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.1996,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330193735042058e-05,
+      "loss": 4.2071,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329356778171379e-05,
+      "loss": 4.2019,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328518183420327e-05,
+      "loss": 4.2057,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327679588669275e-05,
+      "loss": 4.2045,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326840993918223e-05,
+      "loss": 4.2012,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32600239916717e-05,
+      "loss": 4.2125,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325165442296492e-05,
+      "loss": 4.1965,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32432684754544e-05,
+      "loss": 4.1929,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3234882527943885e-05,
+      "loss": 4.1955,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322649658043336e-05,
+      "loss": 4.1959,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321811063292284e-05,
+      "loss": 4.2146,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3209741064216054e-05,
+      "loss": 4.1951,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320135511670553e-05,
+      "loss": 4.2024,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319296916919501e-05,
+      "loss": 4.1818,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318458322168449e-05,
+      "loss": 4.2041,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317619727417397e-05,
+      "loss": 4.1917,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3167827705467176e-05,
+      "loss": 4.1921,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3159441757956656e-05,
+      "loss": 4.2067,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151055810446136e-05,
+      "loss": 4.1965,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314266986293562e-05,
+      "loss": 4.2019,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31342839154251e-05,
+      "loss": 4.2084,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312589796791458e-05,
+      "loss": 4.2016,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311751202040406e-05,
+      "loss": 4.2113,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310912607289354e-05,
+      "loss": 4.1952,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310074012538302e-05,
+      "loss": 4.1859,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309237055667623e-05,
+      "loss": 4.1971,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308398460916571e-05,
+      "loss": 4.2061,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307559866165519e-05,
+      "loss": 4.1949,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306721271414467e-05,
+      "loss": 4.1935,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305884314543788e-05,
+      "loss": 4.1796,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305045719792736e-05,
+      "loss": 4.1922,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304207125041684e-05,
+      "loss": 4.1847,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303368530290632e-05,
+      "loss": 4.1969,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025315734199536e-05,
+      "loss": 4.198,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3016929786689016e-05,
+      "loss": 4.2035,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008560217982226e-05,
+      "loss": 4.1935,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000174270471705e-05,
+      "loss": 4.1946,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2991788322961185e-05,
+      "loss": 4.1812,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983402375450665e-05,
+      "loss": 4.1865,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975016427940145e-05,
+      "loss": 4.1905,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2966646859233354e-05,
+      "loss": 4.1873,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.184,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949874964212314e-05,
+      "loss": 4.1984,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2941489016701794e-05,
+      "loss": 4.1849,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2933103069191274e-05,
+      "loss": 4.2072,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2924717121680754e-05,
+      "loss": 4.1826,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291633117417024e-05,
+      "loss": 4.1992,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2907945226659714e-05,
+      "loss": 4.1884,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289957565795293e-05,
+      "loss": 4.1881,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289118971044241e-05,
+      "loss": 4.2054,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288282014173562e-05,
+      "loss": 4.1816,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28744341942251e-05,
+      "loss": 4.2024,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286604824671458e-05,
+      "loss": 4.1794,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285766229920406e-05,
+      "loss": 4.1779,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284927635169354e-05,
+      "loss": 4.1783,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284090678298675e-05,
+      "loss": 4.1908,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283252083547623e-05,
+      "loss": 4.1889,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282413488796571e-05,
+      "loss": 4.1801,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281574894045519e-05,
+      "loss": 4.1682,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280736299294467e-05,
+      "loss": 4.1809,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279897704543415e-05,
+      "loss": 4.1951,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279059109792363e-05,
+      "loss": 4.1879,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278220515041311e-05,
+      "loss": 4.1925,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277383558170632e-05,
+      "loss": 4.1901,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27654496341958e-05,
+      "loss": 4.189,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275706368668528e-05,
+      "loss": 4.1905,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274867773917476e-05,
+      "loss": 4.1819,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27403245492717e-05,
+      "loss": 4.1783,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273193860176118e-05,
+      "loss": 4.1858,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272355265425066e-05,
+      "loss": 4.1777,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271516670674014e-05,
+      "loss": 4.184,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270678075922962e-05,
+      "loss": 4.1842,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26983948117191e-05,
+      "loss": 4.1842,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269000886420858e-05,
+      "loss": 4.1806,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268162291669806e-05,
+      "loss": 4.1746,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267323696918754e-05,
+      "loss": 4.192,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266486740048075e-05,
+      "loss": 4.1778,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265648145297023e-05,
+      "loss": 4.1893,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264809550545971e-05,
+      "loss": 4.1905,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263970955794919e-05,
+      "loss": 4.1848,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26313399892424e-05,
+      "loss": 4.1857,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2622954041731886e-05,
+      "loss": 4.1797,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614568094221366e-05,
+      "loss": 4.1811,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606182146710846e-05,
+      "loss": 4.1761,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597828956807784e-05,
+      "loss": 4.1948,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589443009297264e-05,
+      "loss": 4.1917,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2581057061786744e-05,
+      "loss": 4.1861,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2572671114276224e-05,
+      "loss": 4.1844,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564285166765704e-05,
+      "loss": 4.1792,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255591559805891e-05,
+      "loss": 4.1836,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254752965054839e-05,
+      "loss": 4.1756,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253916008184161e-05,
+      "loss": 4.1711,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253077413433109e-05,
+      "loss": 4.1824,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252238818682057e-05,
+      "loss": 4.1765,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251400223931005e-05,
+      "loss": 4.1653,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250561629179953e-05,
+      "loss": 4.1771,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.172706604003906,
+      "eval_runtime": 535.8869,
+      "eval_samples_per_second": 712.074,
+      "eval_steps_per_second": 22.253,
+      "step": 457914
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249723034428901e-05,
+      "loss": 4.1869,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248884439677849e-05,
+      "loss": 4.1841,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24804748280717e-05,
+      "loss": 4.179,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247208888056118e-05,
+      "loss": 4.1751,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246370293305066e-05,
+      "loss": 4.1767,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245531698554014e-05,
+      "loss": 4.1657,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244693103802962e-05,
+      "loss": 4.1766,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24385450905191e-05,
+      "loss": 4.1696,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243015914300858e-05,
+      "loss": 4.1725,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242177319549806e-05,
+      "loss": 4.1788,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241338724798754e-05,
+      "loss": 4.1862,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2405001300477024e-05,
+      "loss": 4.1638,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239663173177023e-05,
+      "loss": 4.1799,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238824578425971e-05,
+      "loss": 4.1694,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237985983674919e-05,
+      "loss": 4.1796,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237147388923867e-05,
+      "loss": 4.1654,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236310432053188e-05,
+      "loss": 4.1643,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235471837302136e-05,
+      "loss": 4.1713,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234633242551084e-05,
+      "loss": 4.1702,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2337946478000315e-05,
+      "loss": 4.1794,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2329560530489795e-05,
+      "loss": 4.174,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2321174582979275e-05,
+      "loss": 4.183,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231278863546876e-05,
+      "loss": 4.1775,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230440268795824e-05,
+      "loss": 4.173,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229603311925145e-05,
+      "loss": 4.1681,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228764717174093e-05,
+      "loss": 4.1692,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227926122423041e-05,
+      "loss": 4.1788,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227087527671989e-05,
+      "loss": 4.1684,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226248932920937e-05,
+      "loss": 4.1681,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225410338169885e-05,
+      "loss": 4.1692,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224571743418833e-05,
+      "loss": 4.1611,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223733148667781e-05,
+      "loss": 4.177,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222894553916729e-05,
+      "loss": 4.1753,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.22205759704605e-05,
+      "loss": 4.1753,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221219002294998e-05,
+      "loss": 4.158,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2203820454243196e-05,
+      "loss": 4.1744,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2195434506732676e-05,
+      "loss": 4.1667,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2187048559222156e-05,
+      "loss": 4.1723,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2178662611711635e-05,
+      "loss": 4.1682,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2170276664201115e-05,
+      "loss": 4.1588,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2161890716690595e-05,
+      "loss": 4.1526,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2153504769180075e-05,
+      "loss": 4.1525,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2145118821669555e-05,
+      "loss": 4.1769,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2136749252962764e-05,
+      "loss": 4.1594,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2128363305452244e-05,
+      "loss": 4.1554,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2119977357941724e-05,
+      "loss": 4.1678,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211160778923493e-05,
+      "loss": 4.1626,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210322184172441e-05,
+      "loss": 4.1722,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209483589421389e-05,
+      "loss": 4.158,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208644994670338e-05,
+      "loss": 4.1628,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207806399919286e-05,
+      "loss": 4.1613,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206967805168233e-05,
+      "loss": 4.1678,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206130848297555e-05,
+      "loss": 4.1581,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.205292253546503e-05,
+      "loss": 4.1565,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204453658795451e-05,
+      "loss": 4.1598,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.203615064044398e-05,
+      "loss": 4.1635,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.202776469293346e-05,
+      "loss": 4.1559,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201937874542294e-05,
+      "loss": 4.1561,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201099279791242e-05,
+      "loss": 4.1678,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20026068504019e-05,
+      "loss": 4.1522,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199423728169512e-05,
+      "loss": 4.1487,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19858513341846e-05,
+      "loss": 4.151,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197748176547781e-05,
+      "loss": 4.1537,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196909581796729e-05,
+      "loss": 4.1705,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196070987045677e-05,
+      "loss": 4.151,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195232392294625e-05,
+      "loss": 4.1571,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194393797543573e-05,
+      "loss": 4.1379,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1935568406728936e-05,
+      "loss": 4.158,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1927182459218416e-05,
+      "loss": 4.1494,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1918796511707896e-05,
+      "loss": 4.1517,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1910410564197376e-05,
+      "loss": 4.1604,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1902024616686856e-05,
+      "loss": 4.1521,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1893638669176336e-05,
+      "loss": 4.1585,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1885252721665816e-05,
+      "loss": 4.1643,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18768667741553e-05,
+      "loss": 4.1595,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186849720544851e-05,
+      "loss": 4.1636,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186011125793799e-05,
+      "loss": 4.1586,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.185172531042747e-05,
+      "loss": 4.1412,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184333936291695e-05,
+      "loss": 4.1531,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.183495341540643e-05,
+      "loss": 4.1595,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182656746789591e-05,
+      "loss": 4.156,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181819789918912e-05,
+      "loss": 4.1526,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18098119516786e-05,
+      "loss": 4.1356,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180142600416808e-05,
+      "loss": 4.151,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179304005665756e-05,
+      "loss": 4.1423,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.178465410914704e-05,
+      "loss": 4.1586,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177626816163652e-05,
+      "loss": 4.1532,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1767882214126e-05,
+      "loss": 4.1613,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1759512645419216e-05,
+      "loss": 4.1483,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1751126697908696e-05,
+      "loss": 4.1514,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174274075039817e-05,
+      "loss": 4.1412,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173435480288765e-05,
+      "loss": 4.1511,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172596885537713e-05,
+      "loss": 4.1429,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171758290786661e-05,
+      "loss": 4.1454,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170921333915982e-05,
+      "loss": 4.1393,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17008273916493e-05,
+      "loss": 4.1531,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1692457822942514e-05,
+      "loss": 4.1425,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168407187543199e-05,
+      "loss": 4.1681,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1675685927921474e-05,
+      "loss": 4.1421,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1667299980410954e-05,
+      "loss": 4.1596,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1658914032900434e-05,
+      "loss": 4.1466,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1650528085389914e-05,
+      "loss": 4.1461,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1642142137879393e-05,
+      "loss": 4.1587,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1633756190368873e-05,
+      "loss": 4.1464,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1625370242858353e-05,
+      "loss": 4.1584,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161698429534783e-05,
+      "loss": 4.1446,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160859834783731e-05,
+      "loss": 4.1324,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160021240032679e-05,
+      "loss": 4.1379,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159184283162e-05,
+      "loss": 4.1501,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158345688410948e-05,
+      "loss": 4.1491,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157507093659896e-05,
+      "loss": 4.1424,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156668498908844e-05,
+      "loss": 4.123,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155829904157792e-05,
+      "loss": 4.14,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154992947287114e-05,
+      "loss": 4.1545,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154154352536062e-05,
+      "loss": 4.1472,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15331575778501e-05,
+      "loss": 4.1524,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152477163033958e-05,
+      "loss": 4.1491,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151638568282906e-05,
+      "loss": 4.1476,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150801611412227e-05,
+      "loss": 4.1488,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1499646545415476e-05,
+      "loss": 4.1444,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1491260597904956e-05,
+      "loss": 4.1365,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1482874650394436e-05,
+      "loss": 4.1438,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1474488702883916e-05,
+      "loss": 4.1394,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1466102755373396e-05,
+      "loss": 4.1422,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1457716807862876e-05,
+      "loss": 4.1433,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1449330860352356e-05,
+      "loss": 4.1438,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1440944912841836e-05,
+      "loss": 4.1388,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1432558965331316e-05,
+      "loss": 4.1375,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1424173017820796e-05,
+      "loss": 4.1491,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1415803449114005e-05,
+      "loss": 4.1368,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140743388040722e-05,
+      "loss": 4.1512,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13990479328967e-05,
+      "loss": 4.1508,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1390661985386174e-05,
+      "loss": 4.144,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382276037875654e-05,
+      "loss": 4.1473,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1373890090365134e-05,
+      "loss": 4.142,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1365504142854614e-05,
+      "loss": 4.1394,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357118195344094e-05,
+      "loss": 4.1365,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134874862663731e-05,
+      "loss": 4.1554,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134036267912679e-05,
+      "loss": 4.1517,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133197673161627e-05,
+      "loss": 4.1499,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132360716290948e-05,
+      "loss": 4.1454,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131522121539896e-05,
+      "loss": 4.1372,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130683526788844e-05,
+      "loss": 4.1423,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129844932037792e-05,
+      "loss": 4.1373,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12900633728674e-05,
+      "loss": 4.1303,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128167742535688e-05,
+      "loss": 4.147,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127329147784636e-05,
+      "loss": 4.136,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126492190913957e-05,
+      "loss": 4.1255,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125653596162905e-05,
+      "loss": 4.1395,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.141369819641113,
+      "eval_runtime": 535.6248,
+      "eval_samples_per_second": 712.422,
+      "eval_steps_per_second": 22.264,
+      "step": 534233
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1248150014118534e-05,
+      "loss": 4.1526,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1239764066608014e-05,
+      "loss": 4.1424,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1231378119097494e-05,
+      "loss": 4.1462,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1222992171586974e-05,
+      "loss": 4.1357,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1214606224076454e-05,
+      "loss": 4.1349,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1206220276565934e-05,
+      "loss": 4.1318,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1197834329055414e-05,
+      "loss": 4.1365,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118944838154489e-05,
+      "loss": 4.1313,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118106243403437e-05,
+      "loss": 4.1339,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117267648652385e-05,
+      "loss": 4.1371,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116429053901333e-05,
+      "loss": 4.1493,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115590459150281e-05,
+      "loss": 4.1297,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114751864399229e-05,
+      "loss": 4.141,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113913269648177e-05,
+      "loss": 4.1278,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1130746748971247e-05,
+      "loss": 4.1423,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122360801460727e-05,
+      "loss": 4.128,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111397485395021e-05,
+      "loss": 4.1282,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110558890643969e-05,
+      "loss": 4.1317,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109720295892917e-05,
+      "loss": 4.1329,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108881701141865e-05,
+      "loss": 4.1399,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108044744271186e-05,
+      "loss": 4.1336,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107206149520134e-05,
+      "loss": 4.1477,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106367554769082e-05,
+      "loss": 4.1355,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10552896001803e-05,
+      "loss": 4.1364,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104692003147351e-05,
+      "loss": 4.1339,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103853408396299e-05,
+      "loss": 4.1322,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103014813645247e-05,
+      "loss": 4.1367,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102177856774568e-05,
+      "loss": 4.1319,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101339262023517e-05,
+      "loss": 4.1323,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100500667272465e-05,
+      "loss": 4.1292,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099662072521413e-05,
+      "loss": 4.1224,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0988251156507336e-05,
+      "loss": 4.1358,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0979865208996816e-05,
+      "loss": 4.1404,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0971479261486296e-05,
+      "loss": 4.1343,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963093313975776e-05,
+      "loss": 4.1232,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0954723745268985e-05,
+      "loss": 4.1357,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946354176562194e-05,
+      "loss": 4.13,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0937968229051674e-05,
+      "loss": 4.1338,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0929582281541154e-05,
+      "loss": 4.1294,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921196334030634e-05,
+      "loss": 4.1264,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091281038652012e-05,
+      "loss": 4.1146,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09044244390096e-05,
+      "loss": 4.1148,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089603849149908e-05,
+      "loss": 4.1359,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0887652543988554e-05,
+      "loss": 4.1228,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087928297528177e-05,
+      "loss": 4.1193,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087089702777125e-05,
+      "loss": 4.1316,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086252745906446e-05,
+      "loss": 4.1217,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085414151155394e-05,
+      "loss": 4.1351,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084575556404342e-05,
+      "loss": 4.1213,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08373696165329e-05,
+      "loss": 4.1289,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082900004782611e-05,
+      "loss": 4.1226,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082061410031559e-05,
+      "loss": 4.1294,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0812228152805074e-05,
+      "loss": 4.1236,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0803842205294554e-05,
+      "loss": 4.1173,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079545625778403e-05,
+      "loss": 4.1262,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078708668907724e-05,
+      "loss": 4.1246,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077870074156672e-05,
+      "loss": 4.1216,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0770314794056196e-05,
+      "loss": 4.1201,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0761928846545676e-05,
+      "loss": 4.1324,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0753542899035156e-05,
+      "loss": 4.1121,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745156951524636e-05,
+      "loss": 4.1167,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0736771004014116e-05,
+      "loss": 4.1138,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0728385056503596e-05,
+      "loss": 4.1206,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0719999108993076e-05,
+      "loss": 4.1316,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0711613161482556e-05,
+      "loss": 4.1155,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0703227213972036e-05,
+      "loss": 4.1211,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069484126646152e-05,
+      "loss": 4.1044,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068647169775473e-05,
+      "loss": 4.1222,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067808575024421e-05,
+      "loss": 4.1127,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066969980273369e-05,
+      "loss": 4.1185,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066131385522317e-05,
+      "loss": 4.1232,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065292790771265e-05,
+      "loss": 4.1179,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064455833900586e-05,
+      "loss": 4.12,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063617239149534e-05,
+      "loss": 4.1311,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062778644398482e-05,
+      "loss": 4.1227,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061941687527803e-05,
+      "loss": 4.1274,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061103092776751e-05,
+      "loss": 4.122,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060264498025699e-05,
+      "loss": 4.1077,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594259032746476e-05,
+      "loss": 4.1167,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0585873085235956e-05,
+      "loss": 4.1238,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0577487137725436e-05,
+      "loss": 4.1225,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056910119021491e-05,
+      "loss": 4.1165,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056071524270439e-05,
+      "loss": 4.1037,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0552345673997605e-05,
+      "loss": 4.112,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0543976105290814e-05,
+      "loss": 4.109,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0535590157780294e-05,
+      "loss": 4.1212,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0527204210269774e-05,
+      "loss": 4.117,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0518818262759254e-05,
+      "loss": 4.1241,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0510432315248734e-05,
+      "loss": 4.1178,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0502046367738214e-05,
+      "loss": 4.1125,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0493660420227694e-05,
+      "loss": 4.1103,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0485274472717174e-05,
+      "loss": 4.1147,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047690490401038e-05,
+      "loss": 4.1053,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046851895649986e-05,
+      "loss": 4.1121,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046013300898934e-05,
+      "loss": 4.1084,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045176344028256e-05,
+      "loss": 4.1173,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044337749277203e-05,
+      "loss": 4.1087,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043499154526151e-05,
+      "loss": 4.1337,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042660559775099e-05,
+      "loss": 4.1052,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041821965024047e-05,
+      "loss": 4.1222,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040983370272995e-05,
+      "loss": 4.1168,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040144775521943e-05,
+      "loss": 4.1123,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039307818651265e-05,
+      "loss": 4.1186,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038469223900213e-05,
+      "loss": 4.115,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037630629149161e-05,
+      "loss": 4.122,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036792034398109e-05,
+      "loss": 4.1102,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035953439647057e-05,
+      "loss": 4.1,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035114844896005e-05,
+      "loss": 4.105,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034276250144953e-05,
+      "loss": 4.1148,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033439293274274e-05,
+      "loss": 4.1169,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0326023364035946e-05,
+      "loss": 4.1063,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0317637416525426e-05,
+      "loss": 4.0869,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0309251469014906e-05,
+      "loss": 4.108,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0300865521504386e-05,
+      "loss": 4.1219,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0292479573993866e-05,
+      "loss": 4.1124,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028411000528708e-05,
+      "loss": 4.122,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027572405777656e-05,
+      "loss": 4.115,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026733811026604e-05,
+      "loss": 4.1087,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025895216275552e-05,
+      "loss": 4.1153,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025058259404873e-05,
+      "loss": 4.1114,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024219664653821e-05,
+      "loss": 4.1025,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.023381069902769e-05,
+      "loss": 4.1101,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022542475151717e-05,
+      "loss": 4.1073,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021703880400665e-05,
+      "loss": 4.1074,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020865285649613e-05,
+      "loss": 4.1058,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020026690898561e-05,
+      "loss": 4.1156,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019188096147509e-05,
+      "loss": 4.1063,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0183527771572035e-05,
+      "loss": 4.1005,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175141824061515e-05,
+      "loss": 4.1193,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0166755876550995e-05,
+      "loss": 4.0989,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0158369929040475e-05,
+      "loss": 4.1167,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015001673913741e-05,
+      "loss": 4.1195,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014163079162689e-05,
+      "loss": 4.1122,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013324484411637e-05,
+      "loss": 4.1106,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012485889660585e-05,
+      "loss": 4.1077,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011647294909533e-05,
+      "loss": 4.1073,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010808700158481e-05,
+      "loss": 4.1051,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009970105407429e-05,
+      "loss": 4.1232,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009131510656377e-05,
+      "loss": 4.1156,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008292915905325e-05,
+      "loss": 4.118,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007455959034647e-05,
+      "loss": 4.1096,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006617364283595e-05,
+      "loss": 4.1051,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005778769532543e-05,
+      "loss": 4.1105,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004940174781491e-05,
+      "loss": 4.1011,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004101580030439e-05,
+      "loss": 4.1015,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003262985279387e-05,
+      "loss": 4.1123,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002424390528334e-05,
+      "loss": 4.1048,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001585795777282e-05,
+      "loss": 4.0919,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000748838906604e-05,
+      "loss": 4.1119,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.116526126861572,
+      "eval_runtime": 539.0275,
+      "eval_samples_per_second": 707.925,
+      "eval_steps_per_second": 22.123,
+      "step": 610552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999911882035925e-05,
+      "loss": 4.1073,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999073287284873e-05,
+      "loss": 4.1177,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9982346925338207e-05,
+      "loss": 4.112,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997396097782769e-05,
+      "loss": 4.1022,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9965575030317167e-05,
+      "loss": 4.1025,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9957189082806646e-05,
+      "loss": 4.0993,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9948803135296126e-05,
+      "loss": 4.1036,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9940417187785606e-05,
+      "loss": 4.0977,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9932047619078815e-05,
+      "loss": 4.0997,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9923661671568295e-05,
+      "loss": 4.1068,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991529210286151e-05,
+      "loss": 4.1144,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9906906155350984e-05,
+      "loss": 4.0997,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9898520207840464e-05,
+      "loss": 4.1078,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9890134260329944e-05,
+      "loss": 4.0955,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988174831281943e-05,
+      "loss": 4.116,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987336236530891e-05,
+      "loss": 4.0925,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986499279660212e-05,
+      "loss": 4.0968,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98566068490916e-05,
+      "loss": 4.1013,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984822090158108e-05,
+      "loss": 4.1019,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983983495407056e-05,
+      "loss": 4.1074,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983144900656004e-05,
+      "loss": 4.1014,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.982306305904952e-05,
+      "loss": 4.1166,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9814677111539e-05,
+      "loss": 4.1044,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.980629116402848e-05,
+      "loss": 4.106,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979792159532169e-05,
+      "loss": 4.0993,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.97895520266149e-05,
+      "loss": 4.107,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781166079104385e-05,
+      "loss": 4.1047,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9772780131593865e-05,
+      "loss": 4.1005,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9764394184083345e-05,
+      "loss": 4.1039,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9756008236572825e-05,
+      "loss": 4.0996,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9747622289062305e-05,
+      "loss": 4.0889,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9739236341551785e-05,
+      "loss": 4.1064,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9730866772844994e-05,
+      "loss": 4.1091,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9722480825334474e-05,
+      "loss": 4.1027,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9714094877823954e-05,
+      "loss": 4.0924,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9705708930313434e-05,
+      "loss": 4.1068,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969733936160664e-05,
+      "loss": 4.0987,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968895341409612e-05,
+      "loss": 4.1015,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96805674665856e-05,
+      "loss": 4.1032,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967218151907508e-05,
+      "loss": 4.0924,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966379557156457e-05,
+      "loss": 4.0881,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965540962405405e-05,
+      "loss": 4.0834,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964702367654353e-05,
+      "loss": 4.1068,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9638637729033e-05,
+      "loss": 4.0934,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963025178152248e-05,
+      "loss": 4.0881,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962186583401196e-05,
+      "loss": 4.1041,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961347988650144e-05,
+      "loss": 4.087,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960509393899092e-05,
+      "loss": 4.1093,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959672437028413e-05,
+      "loss": 4.0943,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958833842277361e-05,
+      "loss": 4.1005,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957996885406682e-05,
+      "loss": 4.0951,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957158290655631e-05,
+      "loss": 4.0947,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956319695904579e-05,
+      "loss": 4.0982,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955481101153527e-05,
+      "loss": 4.0824,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9546441442828476e-05,
+      "loss": 4.0976,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9538055495317956e-05,
+      "loss": 4.0924,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9529669547807436e-05,
+      "loss": 4.0908,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9521283600296916e-05,
+      "loss": 4.0929,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9512897652786396e-05,
+      "loss": 4.1025,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9504528084079605e-05,
+      "loss": 4.0799,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9496142136569085e-05,
+      "loss": 4.0904,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9487756189058565e-05,
+      "loss": 4.0868,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9479370241548045e-05,
+      "loss": 4.0912,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9470984294037525e-05,
+      "loss": 4.0985,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9462598346527005e-05,
+      "loss": 4.088,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9454212399016485e-05,
+      "loss": 4.0973,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944582645150597e-05,
+      "loss": 4.0693,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943747326160291e-05,
+      "loss": 4.0973,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942908731409239e-05,
+      "loss": 4.0829,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942070136658187e-05,
+      "loss": 4.0865,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941231541907135e-05,
+      "loss": 4.0922,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940392947156083e-05,
+      "loss": 4.093,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939555990285404e-05,
+      "loss": 4.09,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938717395534352e-05,
+      "loss": 4.1049,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9378788007833e-05,
+      "loss": 4.0911,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937040206032248e-05,
+      "loss": 4.0995,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.936201611281196e-05,
+      "loss": 4.0956,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9353646544105174e-05,
+      "loss": 4.0818,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9345260596594654e-05,
+      "loss": 4.0835,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9336874649084134e-05,
+      "loss": 4.0937,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9328488701573614e-05,
+      "loss": 4.0973,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9320102754063094e-05,
+      "loss": 4.0932,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9311716806552574e-05,
+      "loss": 4.071,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9303330859042054e-05,
+      "loss": 4.0817,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9294944911531534e-05,
+      "loss": 4.0829,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928657534282474e-05,
+      "loss": 4.0899,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927818939531422e-05,
+      "loss": 4.0915,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92698034478037e-05,
+      "loss": 4.0941,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926141750029318e-05,
+      "loss": 4.0878,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925303155278266e-05,
+      "loss": 4.0863,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924464560527214e-05,
+      "loss": 4.0819,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923625965776162e-05,
+      "loss": 4.0844,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.92278737102511e-05,
+      "loss": 4.074,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921950414154431e-05,
+      "loss": 4.0906,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921111819403379e-05,
+      "loss": 4.082,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920273224652327e-05,
+      "loss": 4.0864,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919436267781648e-05,
+      "loss": 4.0809,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918597673030596e-05,
+      "loss": 4.103,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917759078279544e-05,
+      "loss": 4.0792,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916920483528492e-05,
+      "loss": 4.0945,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91608188877744e-05,
+      "loss": 4.0908,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.915243294026388e-05,
+      "loss": 4.0862,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9144063371557096e-05,
+      "loss": 4.0878,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913569380285031e-05,
+      "loss": 4.0903,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9127307855339786e-05,
+      "loss": 4.0933,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9118921907829266e-05,
+      "loss": 4.081,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9110535960318745e-05,
+      "loss": 4.0746,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9102150012808225e-05,
+      "loss": 4.0779,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9093764065297705e-05,
+      "loss": 4.0862,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9085378117787185e-05,
+      "loss": 4.0857,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9076992170276665e-05,
+      "loss": 4.0841,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9068606222766145e-05,
+      "loss": 4.0595,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9060220275255625e-05,
+      "loss": 4.0836,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9051834327745105e-05,
+      "loss": 4.089,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9043448380234585e-05,
+      "loss": 4.0857,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90350788115278e-05,
+      "loss": 4.0932,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902670924282101e-05,
+      "loss": 4.0886,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901832329531049e-05,
+      "loss": 4.0783,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900993734779997e-05,
+      "loss": 4.0901,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900155140028945e-05,
+      "loss": 4.0875,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.899316545277893e-05,
+      "loss": 4.0742,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898477950526841e-05,
+      "loss": 4.0842,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.897639355775789e-05,
+      "loss": 4.0796,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.896800761024737e-05,
+      "loss": 4.0793,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895963804154058e-05,
+      "loss": 4.0798,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895125209403006e-05,
+      "loss": 4.0849,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.894286614651954e-05,
+      "loss": 4.0854,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.893448019900902e-05,
+      "loss": 4.0719,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89260942514985e-05,
+      "loss": 4.0884,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8917724682791715e-05,
+      "loss": 4.075,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8909338735281195e-05,
+      "loss": 4.0891,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8900969166574404e-05,
+      "loss": 4.0967,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8892583219063884e-05,
+      "loss": 4.0819,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8884197271553364e-05,
+      "loss": 4.0853,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8875811324042843e-05,
+      "loss": 4.0782,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886742537653232e-05,
+      "loss": 4.0819,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8859039429021797e-05,
+      "loss": 4.0787,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8850653481511277e-05,
+      "loss": 4.094,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8842267534000757e-05,
+      "loss": 4.0946,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8833881586490236e-05,
+      "loss": 4.0857,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882551201778345e-05,
+      "loss": 4.0825,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881712607027293e-05,
+      "loss": 4.0842,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880874012276241e-05,
+      "loss": 4.0808,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880035417525189e-05,
+      "loss": 4.0781,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.87919846065451e-05,
+      "loss": 4.0741,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878359865903458e-05,
+      "loss": 4.0847,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877521271152406e-05,
+      "loss": 4.0792,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876682676401354e-05,
+      "loss": 4.0689,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.875845719530675e-05,
+      "loss": 4.0787,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.097525119781494,
+      "eval_runtime": 536.3073,
+      "eval_samples_per_second": 711.516,
+      "eval_steps_per_second": 22.235,
+      "step": 686871
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8750087626599966e-05,
+      "loss": 4.0873,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874170167908944e-05,
+      "loss": 4.0887,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8733315731578926e-05,
+      "loss": 4.0877,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8724929784068406e-05,
+      "loss": 4.0718,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8716543836557886e-05,
+      "loss": 4.0823,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8708157889047366e-05,
+      "loss": 4.0729,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8699771941536846e-05,
+      "loss": 4.0748,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8691385994026326e-05,
+      "loss": 4.0741,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8683000046515806e-05,
+      "loss": 4.0713,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8674614099005286e-05,
+      "loss": 4.0843,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8666244530298495e-05,
+      "loss": 4.0847,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8657858582787975e-05,
+      "loss": 4.0726,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8649472635277455e-05,
+      "loss": 4.0854,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8641086687766935e-05,
+      "loss": 4.0689,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8632700740256415e-05,
+      "loss": 4.0833,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8624314792745895e-05,
+      "loss": 4.0671,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8615928845235375e-05,
+      "loss": 4.0725,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8607542897724855e-05,
+      "loss": 4.0773,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599156950214334e-05,
+      "loss": 4.0742,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8590771002703814e-05,
+      "loss": 4.0846,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8582385055193294e-05,
+      "loss": 4.0696,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8573999107682774e-05,
+      "loss": 4.0948,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8565613160172254e-05,
+      "loss": 4.074,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8557227212661734e-05,
+      "loss": 4.0794,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8548841265151214e-05,
+      "loss": 4.0762,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854047169644442e-05,
+      "loss": 4.0813,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85320857489339e-05,
+      "loss": 4.0782,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852369980142338e-05,
+      "loss": 4.0714,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851531385391286e-05,
+      "loss": 4.0775,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850692790640234e-05,
+      "loss": 4.0764,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849855833769556e-05,
+      "loss": 4.0583,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849017239018504e-05,
+      "loss": 4.0821,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.848178644267452e-05,
+      "loss": 4.0844,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8473400495164e-05,
+      "loss": 4.0787,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846501454765348e-05,
+      "loss": 4.067,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845664497894669e-05,
+      "loss": 4.0818,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84482754102399e-05,
+      "loss": 4.0754,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843988946272938e-05,
+      "loss": 4.0747,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843150351521886e-05,
+      "loss": 4.0756,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842311756770834e-05,
+      "loss": 4.0651,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841473162019782e-05,
+      "loss": 4.0651,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84063456726873e-05,
+      "loss": 4.0577,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839795972517678e-05,
+      "loss": 4.0804,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838957377766626e-05,
+      "loss": 4.069,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8381187830155743e-05,
+      "loss": 4.0627,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8372801882645223e-05,
+      "loss": 4.0759,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836443231393843e-05,
+      "loss": 4.0622,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835604636642791e-05,
+      "loss": 4.0797,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834766041891739e-05,
+      "loss": 4.0736,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8339274471406866e-05,
+      "loss": 4.0725,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8330888523896346e-05,
+      "loss": 4.0711,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8322502576385826e-05,
+      "loss": 4.0652,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314116628875305e-05,
+      "loss": 4.0744,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8305747060168515e-05,
+      "loss": 4.0559,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8297361112657995e-05,
+      "loss": 4.0747,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828897516514748e-05,
+      "loss": 4.0639,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828058921763696e-05,
+      "loss": 4.0681,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827220327012644e-05,
+      "loss": 4.0638,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8263850080223386e-05,
+      "loss": 4.0773,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8255464132712866e-05,
+      "loss": 4.0616,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824707818520234e-05,
+      "loss": 4.0594,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823869223769182e-05,
+      "loss": 4.0657,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82303062901813e-05,
+      "loss": 4.06,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8221936721474515e-05,
+      "loss": 4.0727,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821355077396399e-05,
+      "loss": 4.0695,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820516482645347e-05,
+      "loss": 4.0707,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819677887894295e-05,
+      "loss": 4.046,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8188392931432435e-05,
+      "loss": 4.0711,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8180006983921915e-05,
+      "loss": 4.0609,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171621036411395e-05,
+      "loss": 4.0612,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8163235088900875e-05,
+      "loss": 4.0641,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154865520194084e-05,
+      "loss": 4.0667,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8146479572683564e-05,
+      "loss": 4.064,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8138093625173044e-05,
+      "loss": 4.082,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8129707677662524e-05,
+      "loss": 4.0669,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812133810895573e-05,
+      "loss": 4.0731,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811295216144521e-05,
+      "loss": 4.0679,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810456621393469e-05,
+      "loss": 4.0589,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809618026642417e-05,
+      "loss": 4.059,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808781069771739e-05,
+      "loss": 4.0664,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807942475020687e-05,
+      "loss": 4.07,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807103880269635e-05,
+      "loss": 4.0744,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806265285518583e-05,
+      "loss": 4.0451,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805426690767531e-05,
+      "loss": 4.0538,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804588096016479e-05,
+      "loss": 4.0586,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8037527770261727e-05,
+      "loss": 4.0665,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8029141822751207e-05,
+      "loss": 4.0641,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8020755875240687e-05,
+      "loss": 4.0733,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8012369927730166e-05,
+      "loss": 4.0632,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8003983980219646e-05,
+      "loss": 4.0622,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7995598032709126e-05,
+      "loss": 4.0586,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7987212085198606e-05,
+      "loss": 4.0606,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7978826137688086e-05,
+      "loss": 4.0494,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79704565689813e-05,
+      "loss": 4.0629,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796207062147078e-05,
+      "loss": 4.0618,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795368467396026e-05,
+      "loss": 4.0597,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794529872644974e-05,
+      "loss": 4.0589,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793691277893922e-05,
+      "loss": 4.0735,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792854321023243e-05,
+      "loss": 4.0574,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792015726272191e-05,
+      "loss": 4.0697,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.791177131521139e-05,
+      "loss": 4.0654,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.790338536770087e-05,
+      "loss": 4.0653,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789501579899408e-05,
+      "loss": 4.0591,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7886646230287296e-05,
+      "loss": 4.0666,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7878260282776776e-05,
+      "loss": 4.0691,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7869874335266256e-05,
+      "loss": 4.0557,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7861488387755736e-05,
+      "loss": 4.0534,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7853102440245216e-05,
+      "loss": 4.0514,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7844716492734696e-05,
+      "loss": 4.0591,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7836346924027905e-05,
+      "loss": 4.0627,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7827960976517385e-05,
+      "loss": 4.0597,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7819575029006865e-05,
+      "loss": 4.0414,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7811189081496345e-05,
+      "loss": 4.0543,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7802803133985825e-05,
+      "loss": 4.0621,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.77944171864753e-05,
+      "loss": 4.0651,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778603123896478e-05,
+      "loss": 4.0722,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7777645291454264e-05,
+      "loss": 4.0629,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7769275722747474e-05,
+      "loss": 4.0566,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7760889775236954e-05,
+      "loss": 4.0619,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7752503827726433e-05,
+      "loss": 4.066,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774413425901965e-05,
+      "loss": 4.0533,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773576469031286e-05,
+      "loss": 4.0547,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772737874280234e-05,
+      "loss": 4.0603,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771899279529182e-05,
+      "loss": 4.0493,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.77106068477813e-05,
+      "loss": 4.0612,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770222090027077e-05,
+      "loss": 4.058,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769383495276025e-05,
+      "loss": 4.0627,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768544900524973e-05,
+      "loss": 4.0487,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767706305773921e-05,
+      "loss": 4.0623,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76686771102287e-05,
+      "loss": 4.0518,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766029116271818e-05,
+      "loss": 4.0668,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765190521520766e-05,
+      "loss": 4.0715,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764351926769714e-05,
+      "loss": 4.0582,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763514969899035e-05,
+      "loss": 4.0602,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762676375147983e-05,
+      "loss": 4.0608,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7618394182773036e-05,
+      "loss": 4.0579,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7610008235262516e-05,
+      "loss": 4.056,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7601622287751996e-05,
+      "loss": 4.0696,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7593236340241476e-05,
+      "loss": 4.0656,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7584850392730956e-05,
+      "loss": 4.0654,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7576464445220436e-05,
+      "loss": 4.0606,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756809487651365e-05,
+      "loss": 4.0654,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755970892900313e-05,
+      "loss": 4.0534,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755132298149261e-05,
+      "loss": 4.053,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754293703398209e-05,
+      "loss": 4.0523,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753455108647157e-05,
+      "loss": 4.062,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752616513896105e-05,
+      "loss": 4.056,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751777919145053e-05,
+      "loss": 4.0411,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.750940962274374e-05,
+      "loss": 4.0592,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.081372261047363,
+      "eval_runtime": 537.7049,
+      "eval_samples_per_second": 709.666,
+      "eval_steps_per_second": 22.178,
+      "step": 763190
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750102367523322e-05,
+      "loss": 4.0626,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74926377277227e-05,
+      "loss": 4.0633,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748425178021218e-05,
+      "loss": 4.0663,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7475865832701654e-05,
+      "loss": 4.0463,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7467479885191134e-05,
+      "loss": 4.0626,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745911031648435e-05,
+      "loss": 4.0527,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7450724368973836e-05,
+      "loss": 4.0513,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744233842146331e-05,
+      "loss": 4.0482,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743395247395279e-05,
+      "loss": 4.0506,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742556652644227e-05,
+      "loss": 4.0612,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7417196957735485e-05,
+      "loss": 4.0593,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740881101022496e-05,
+      "loss": 4.0508,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740042506271444e-05,
+      "loss": 4.0612,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739203911520392e-05,
+      "loss": 4.0451,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.73836531676934e-05,
+      "loss": 4.0628,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737526722018288e-05,
+      "loss": 4.044,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736689765147609e-05,
+      "loss": 4.0484,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7358511703965574e-05,
+      "loss": 4.0551,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7350125756455054e-05,
+      "loss": 4.0539,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7341739808944534e-05,
+      "loss": 4.0586,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7333353861434014e-05,
+      "loss": 4.0453,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7324967913923494e-05,
+      "loss": 4.0726,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7316581966412974e-05,
+      "loss": 4.0533,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730821239770618e-05,
+      "loss": 4.0587,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729982645019566e-05,
+      "loss": 4.0511,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729144050268514e-05,
+      "loss": 4.0615,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728305455517462e-05,
+      "loss": 4.0546,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.72746686076641e-05,
+      "loss": 4.0488,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726628266015358e-05,
+      "loss": 4.0557,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725789671264306e-05,
+      "loss": 4.0537,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724951076513254e-05,
+      "loss": 4.0376,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724114119642576e-05,
+      "loss": 4.0572,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.723277162771897e-05,
+      "loss": 4.0599,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722438568020845e-05,
+      "loss": 4.0553,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721599973269793e-05,
+      "loss": 4.0446,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720761378518741e-05,
+      "loss": 4.0635,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7199244216480616e-05,
+      "loss": 4.0501,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7190858268970096e-05,
+      "loss": 4.0574,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7182472321459576e-05,
+      "loss": 4.0527,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7174086373949056e-05,
+      "loss": 4.0433,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7165700426438536e-05,
+      "loss": 4.0465,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7157314478928016e-05,
+      "loss": 4.0352,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7148928531417496e-05,
+      "loss": 4.0553,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7140542583906976e-05,
+      "loss": 4.0463,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7132156636396456e-05,
+      "loss": 4.0408,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7123770688885936e-05,
+      "loss": 4.0531,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7115401120179145e-05,
+      "loss": 4.0407,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7107015172668625e-05,
+      "loss": 4.0567,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7098629225158105e-05,
+      "loss": 4.0506,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7090243277647585e-05,
+      "loss": 4.0473,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7081857330137065e-05,
+      "loss": 4.0491,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7073471382626545e-05,
+      "loss": 4.0449,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7065085435116025e-05,
+      "loss": 4.0499,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7056699487605505e-05,
+      "loss": 4.0358,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7048329918898714e-05,
+      "loss": 4.0501,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703996035019193e-05,
+      "loss": 4.0451,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703157440268141e-05,
+      "loss": 4.0452,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702318845517089e-05,
+      "loss": 4.0428,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701480250766037e-05,
+      "loss": 4.0527,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700643293895358e-05,
+      "loss": 4.0409,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699804699144306e-05,
+      "loss": 4.0382,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698966104393254e-05,
+      "loss": 4.0447,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698127509642202e-05,
+      "loss": 4.0373,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69728891489115e-05,
+      "loss": 4.049,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696450320140098e-05,
+      "loss": 4.0475,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695611725389046e-05,
+      "loss": 4.0473,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694773130637994e-05,
+      "loss": 4.0257,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693934535886942e-05,
+      "loss": 4.0502,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6930992168966363e-05,
+      "loss": 4.0385,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6922606221455843e-05,
+      "loss": 4.0394,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691422027394532e-05,
+      "loss": 4.0395,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69058343264348e-05,
+      "loss": 4.0486,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689744837892428e-05,
+      "loss": 4.0396,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688906243141376e-05,
+      "loss": 4.0603,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688069286270697e-05,
+      "loss": 4.0473,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687230691519645e-05,
+      "loss": 4.0476,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686392096768593e-05,
+      "loss": 4.0472,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685553502017541e-05,
+      "loss": 4.0365,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.684714907266489e-05,
+      "loss": 4.0383,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683876312515437e-05,
+      "loss": 4.0466,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683039355644758e-05,
+      "loss": 4.0463,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682200760893707e-05,
+      "loss": 4.0518,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681363804023028e-05,
+      "loss": 4.0248,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680525209271976e-05,
+      "loss": 4.0324,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679686614520924e-05,
+      "loss": 4.0347,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678848019769872e-05,
+      "loss": 4.0475,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.67800942501882e-05,
+      "loss": 4.0402,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677170830267768e-05,
+      "loss": 4.0501,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676332235516715e-05,
+      "loss": 4.0438,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675493640765663e-05,
+      "loss": 4.0431,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674655046014611e-05,
+      "loss": 4.0331,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673816451263559e-05,
+      "loss": 4.041,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672977856512507e-05,
+      "loss": 4.031,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672139261761455e-05,
+      "loss": 4.0394,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6713006670104037e-05,
+      "loss": 4.0382,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6704637101397246e-05,
+      "loss": 4.04,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6696251153886726e-05,
+      "loss": 4.0402,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6687865206376206e-05,
+      "loss": 4.0489,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6679479258865685e-05,
+      "loss": 4.042,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6671093311355165e-05,
+      "loss": 4.0473,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6662723742648375e-05,
+      "loss": 4.0405,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6654354173941584e-05,
+      "loss": 4.0445,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6645968226431064e-05,
+      "loss": 4.0402,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6637582278920544e-05,
+      "loss": 4.0476,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6629196331410023e-05,
+      "loss": 4.0455,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6620810383899503e-05,
+      "loss": 4.0332,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6612424436388983e-05,
+      "loss": 4.0322,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660403848887847e-05,
+      "loss": 4.0282,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.659566892017168e-05,
+      "loss": 4.0412,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658728297266116e-05,
+      "loss": 4.041,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657889702515064e-05,
+      "loss": 4.0445,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657051107764012e-05,
+      "loss": 4.0185,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65621251301296e-05,
+      "loss": 4.0268,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.655373918261908e-05,
+      "loss": 4.0436,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.654535323510856e-05,
+      "loss": 4.0445,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653696728759804e-05,
+      "loss": 4.0517,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652859771889125e-05,
+      "loss": 4.0454,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652021177138073e-05,
+      "loss": 4.031,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651182582387021e-05,
+      "loss": 4.0421,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650343987635969e-05,
+      "loss": 4.0413,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6495070307652904e-05,
+      "loss": 4.0325,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6486684360142384e-05,
+      "loss": 4.0318,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647831479143559e-05,
+      "loss": 4.0386,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646992884392507e-05,
+      "loss": 4.0337,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646154289641455e-05,
+      "loss": 4.0379,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645315694890403e-05,
+      "loss": 4.0343,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644477100139351e-05,
+      "loss": 4.0452,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643640143268672e-05,
+      "loss": 4.0298,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64280154851762e-05,
+      "loss": 4.037,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641962953766568e-05,
+      "loss": 4.0367,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641124359015516e-05,
+      "loss": 4.0485,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640285764264464e-05,
+      "loss": 4.0473,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639447169513412e-05,
+      "loss": 4.0361,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63860857476236e-05,
+      "loss": 4.0392,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637769980011308e-05,
+      "loss": 4.0414,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636933023140629e-05,
+      "loss": 4.0366,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636094428389577e-05,
+      "loss": 4.034,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635255833638525e-05,
+      "loss": 4.048,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634417238887473e-05,
+      "loss": 4.0446,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.633580282016794e-05,
+      "loss": 4.0446,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632741687265742e-05,
+      "loss": 4.0395,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63190309251469e-05,
+      "loss": 4.0432,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631064497763638e-05,
+      "loss": 4.0297,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630225903012586e-05,
+      "loss": 4.033,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6293873082615346e-05,
+      "loss": 4.0339,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6285487135104826e-05,
+      "loss": 4.0393,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6277101187594306e-05,
+      "loss": 4.0349,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6268731618887515e-05,
+      "loss": 4.023,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6260362050180724e-05,
+      "loss": 4.0379,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.067577362060547,
+      "eval_runtime": 550.3917,
+      "eval_samples_per_second": 693.308,
+      "eval_steps_per_second": 21.666,
+      "step": 839509
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.399587165752399e+17,
+  "trial_name": null,
+  "trial_params": null
+}