diff --git "a/checkpoint-1984320/trainer_state.json" "b/checkpoint-1984320/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1984320/trainer_state.json"
@@ -0,0 +1,23483 @@
+{
+  "best_metric": 3.985201120376587,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-reconstruction/lstm/3/checkpoints/checkpoint-1984320",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1984320,
+  "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": 10.8204,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5463,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0572,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.992,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9466,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8848,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7333,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6274,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5278,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4492,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3879,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3324,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2709,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2013,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1486,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.086,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0514,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 6.0111,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9594,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.9279,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8891,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8571,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.828,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7935,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.776,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979036769104073e-05,
+      "loss": 5.7382,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.978198174353021e-05,
+      "loss": 5.7199,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6947,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6663,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975684027980238e-05,
+      "loss": 5.6387,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974845433229186e-05,
+      "loss": 5.6242,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974006838478134e-05,
+      "loss": 5.612,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973168243727082e-05,
+      "loss": 5.5866,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5653,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714926921053515e-05,
+      "loss": 5.5573,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706540973542995e-05,
+      "loss": 5.5454,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698155026032475e-05,
+      "loss": 5.5198,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689769078521955e-05,
+      "loss": 5.4864,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9681399509815164e-05,
+      "loss": 5.4823,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9673013562304644e-05,
+      "loss": 5.4489,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9664627614794124e-05,
+      "loss": 5.4544,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9656241667283604e-05,
+      "loss": 5.4369,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4347,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4063,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.4035,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962271425604525e-05,
+      "loss": 5.3896,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961432830853473e-05,
+      "loss": 5.3818,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3775,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.343,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3416,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.3449,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3357,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3099,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.297,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.2922,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2822,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2933,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2534,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2642,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2538,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2382,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2402,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2133,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.2084,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.2065,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.218,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1881,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1832,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1637,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1706,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1721,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1709,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1466,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1601,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1599,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1345,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1218,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.114,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.1038,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.1139,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0934,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.1094,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.1001,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0839,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 5.0688,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0705,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.0713,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.0748,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0549,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0539,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0498,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0342,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0315,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.03,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0216,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0207,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.0093,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 5.0018,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 5.0083,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 5.003,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 4.9973,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9909,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9808,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9752,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9725,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9504,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9604,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9562,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.959,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.9439,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.9504,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906945464479944e-05,
+      "loss": 4.9265,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906106869728892e-05,
+      "loss": 4.9478,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052682749778396e-05,
+      "loss": 4.9173,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9044296802267876e-05,
+      "loss": 4.9268,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9273,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.9219,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019155338540045e-05,
+      "loss": 4.9183,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010769391029525e-05,
+      "loss": 4.9056,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002383443519005e-05,
+      "loss": 4.9103,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8993997496008485e-05,
+      "loss": 4.9138,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985611548497965e-05,
+      "loss": 4.8967,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977225600987445e-05,
+      "loss": 4.8921,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896885603228066e-05,
+      "loss": 4.8895,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896047008477014e-05,
+      "loss": 4.9014,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8686,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8732,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8641,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8732,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8603,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8786,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.8672,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8588,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.8499,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766433672724e-05,
+      "loss": 4.8562,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886825741976188e-05,
+      "loss": 4.8531,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859871472251365e-05,
+      "loss": 4.8521,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851485524740845e-05,
+      "loss": 4.8394,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843099577230325e-05,
+      "loss": 4.8502,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834730008523534e-05,
+      "loss": 4.8381,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826344061013014e-05,
+      "loss": 4.8359,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8817958113502494e-05,
+      "loss": 4.8194,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.8344,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8183,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879281664977466e-05,
+      "loss": 4.8001,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878443070226414e-05,
+      "loss": 4.8267,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.8134,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.8117,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.8098,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87509032910258e-05,
+      "loss": 4.7957,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.76328706741333,
+      "eval_runtime": 292.912,
+      "eval_samples_per_second": 1302.75,
+      "eval_steps_per_second": 40.712,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874251734351528e-05,
+      "loss": 4.7803,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873413139600475e-05,
+      "loss": 4.7865,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.8035,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871737587978745e-05,
+      "loss": 4.7888,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870898993227693e-05,
+      "loss": 4.793,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87006039847664e-05,
+      "loss": 4.7814,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869221803725588e-05,
+      "loss": 4.7816,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7666,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.7698,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.7717,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658690626017536e-05,
+      "loss": 4.7832,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650304678507016e-05,
+      "loss": 4.7748,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8641935109800225e-05,
+      "loss": 4.7599,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633549162289705e-05,
+      "loss": 4.7534,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625163214779185e-05,
+      "loss": 4.7428,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616777267268665e-05,
+      "loss": 4.7446,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608391319758145e-05,
+      "loss": 4.7568,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600005372247625e-05,
+      "loss": 4.7431,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591619424737105e-05,
+      "loss": 4.7391,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583233477226585e-05,
+      "loss": 4.7551,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574863908519794e-05,
+      "loss": 4.7432,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566477961009274e-05,
+      "loss": 4.736,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7376,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.7488,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854133649728145e-05,
+      "loss": 4.7214,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.7251,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.7337,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.7228,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85078090860431e-05,
+      "loss": 4.7081,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849942313853258e-05,
+      "loss": 4.7107,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849103719102206e-05,
+      "loss": 4.7248,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848265124351154e-05,
+      "loss": 4.7158,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847426529600102e-05,
+      "loss": 4.7116,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84658793484905e-05,
+      "loss": 4.7159,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.7116,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.709,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.691,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.696,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.68,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.694,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840721047352432e-05,
+      "loss": 4.69,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83988245260138e-05,
+      "loss": 4.6986,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6796,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838208538860022e-05,
+      "loss": 4.6909,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6812,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.6834,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6892,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6591,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6642,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.6829,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6742,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6642,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6549,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6576,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.6468,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.6732,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.6411,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.6603,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.6497,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6426,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.6466,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823120384862579e-05,
+      "loss": 4.6363,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822281790111527e-05,
+      "loss": 4.6291,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6366,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6525,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.6239,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.6284,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.614,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.6278,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6399,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.6336,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.6211,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81389748048138e-05,
+      "loss": 4.6356,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.6402,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812220290979276e-05,
+      "loss": 4.6196,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.6139,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810544739357545e-05,
+      "loss": 4.6156,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.6071,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808867549855441e-05,
+      "loss": 4.6165,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808028955104389e-05,
+      "loss": 4.6035,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.617,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6154,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.6105,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.5933,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.6017,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.6047,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.6103,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.6037,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800484878105667e-05,
+      "loss": 4.5975,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.5994,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.5878,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5856,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797133774862205e-05,
+      "loss": 4.5948,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796295180111153e-05,
+      "loss": 4.5851,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.5889,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5877,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.5735,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792942438987319e-05,
+      "loss": 4.5867,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.5918,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.5825,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5787,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.57,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5747,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879125083613796e-05,
+      "loss": 4.5746,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870739136103276e-05,
+      "loss": 4.5559,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862369567396485e-05,
+      "loss": 4.5617,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853983619885965e-05,
+      "loss": 4.5648,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845597672375445e-05,
+      "loss": 4.5735,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837211724864925e-05,
+      "loss": 4.5607,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.5651,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.5494,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.5728,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803667934822844e-05,
+      "loss": 4.5506,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795281987312324e-05,
+      "loss": 4.5542,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.5644,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.5597,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777014052358449e-05,
+      "loss": 4.5582,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.5448,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.555,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5614,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.5425,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.5454,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.5422,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5561,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.5286,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.5389,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5273,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767791147977251e-05,
+      "loss": 4.5408,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766952553226199e-05,
+      "loss": 4.5286,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76611559635552e-05,
+      "loss": 4.5497,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765277001604468e-05,
+      "loss": 4.5429,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764438406853416e-05,
+      "loss": 4.5343,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763599812102364e-05,
+      "loss": 4.5265,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762762855231685e-05,
+      "loss": 4.5382,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761924260480633e-05,
+      "loss": 4.5359,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610856657295813e-05,
+      "loss": 4.541,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602470709785293e-05,
+      "loss": 4.5203,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941011410785e-05,
+      "loss": 4.5389,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.5301,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757732924605746e-05,
+      "loss": 4.5318,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756894329854694e-05,
+      "loss": 4.5181,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756057372984015e-05,
+      "loss": 4.5307,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.5188,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5094,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753543226611232e-05,
+      "loss": 4.5262,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75270463186018e-05,
+      "loss": 4.523,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751866037109128e-05,
+      "loss": 4.5198,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751027442358077e-05,
+      "loss": 4.5164,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750188847607025e-05,
+      "loss": 4.5072,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.479551792144775,
+      "eval_runtime": 293.648,
+      "eval_samples_per_second": 1299.484,
+      "eval_steps_per_second": 40.61,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749350252855973e-05,
+      "loss": 4.4923,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748511658104921e-05,
+      "loss": 4.5003,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747673063353868e-05,
+      "loss": 4.5209,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746834468602816e-05,
+      "loss": 4.5089,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745995873851764e-05,
+      "loss": 4.5136,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745157279100712e-05,
+      "loss": 4.4992,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74431868434966e-05,
+      "loss": 4.5068,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743480089598608e-05,
+      "loss": 4.4937,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742641494847556e-05,
+      "loss": 4.4974,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741802900096504e-05,
+      "loss": 4.5044,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740964305345452e-05,
+      "loss": 4.5089,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401257105944e-05,
+      "loss": 4.5105,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392887537237216e-05,
+      "loss": 4.4918,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384501589726696e-05,
+      "loss": 4.4902,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376115642216176e-05,
+      "loss": 4.4886,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367729694705656e-05,
+      "loss": 4.4776,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359343747195135e-05,
+      "loss": 4.495,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350957799684615e-05,
+      "loss": 4.4832,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4853,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.5042,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.487,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4865,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730904444093574e-05,
+      "loss": 4.4901,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730065849342522e-05,
+      "loss": 4.5016,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292272545914704e-05,
+      "loss": 4.4709,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7283886598404184e-05,
+      "loss": 4.478,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275500650893664e-05,
+      "loss": 4.487,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.4817,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4674,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4659,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724197323965531e-05,
+      "loss": 4.4854,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723358729214479e-05,
+      "loss": 4.483,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722520134463427e-05,
+      "loss": 4.4735,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721681539712375e-05,
+      "loss": 4.4804,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.4766,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.473,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.459,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718330436468914e-05,
+      "loss": 4.4702,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717493479598235e-05,
+      "loss": 4.449,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716654884847183e-05,
+      "loss": 4.4668,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715816290096131e-05,
+      "loss": 4.4636,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714977695345079e-05,
+      "loss": 4.4712,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714139100594027e-05,
+      "loss": 4.4588,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713300505842975e-05,
+      "loss": 4.4657,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712461911091923e-05,
+      "loss": 4.463,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711623316340871e-05,
+      "loss": 4.4588,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107863594701916e-05,
+      "loss": 4.4719,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099477647191396e-05,
+      "loss": 4.4427,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091091699680876e-05,
+      "loss": 4.4503,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082705752170356e-05,
+      "loss": 4.4636,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707433618346357e-05,
+      "loss": 4.4602,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706595023595305e-05,
+      "loss": 4.4514,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705756428844253e-05,
+      "loss": 4.4418,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704917834093201e-05,
+      "loss": 4.45,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704080877222522e-05,
+      "loss": 4.4314,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70324228247147e-05,
+      "loss": 4.4624,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702403687720418e-05,
+      "loss": 4.4371,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701565092969366e-05,
+      "loss": 4.4445,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700728136098687e-05,
+      "loss": 4.4453,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699889541347635e-05,
+      "loss": 4.4368,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699050946596583e-05,
+      "loss": 4.442,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698212351845531e-05,
+      "loss": 4.4331,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973753949748525e-05,
+      "loss": 4.4289,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965368002238005e-05,
+      "loss": 4.4313,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956982054727485e-05,
+      "loss": 4.4503,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948596107216965e-05,
+      "loss": 4.4289,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940210159706445e-05,
+      "loss": 4.4291,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931824212195925e-05,
+      "loss": 4.4173,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69234382646854e-05,
+      "loss": 4.4307,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915068695978614e-05,
+      "loss": 4.4414,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690669912727182e-05,
+      "loss": 4.442,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68983131797613e-05,
+      "loss": 4.4267,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688992723225078e-05,
+      "loss": 4.4432,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688154128474026e-05,
+      "loss": 4.4493,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687315533722974e-05,
+      "loss": 4.427,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686476938971923e-05,
+      "loss": 4.4171,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685639982101244e-05,
+      "loss": 4.4307,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684801387350192e-05,
+      "loss": 4.4206,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68396279259914e-05,
+      "loss": 4.4275,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683124197848087e-05,
+      "loss": 4.4153,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682285603097035e-05,
+      "loss": 4.4296,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681447008345983e-05,
+      "loss": 4.4343,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680608413594931e-05,
+      "loss": 4.423,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679769818843879e-05,
+      "loss": 4.4085,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789328619732e-05,
+      "loss": 4.4177,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678094267222148e-05,
+      "loss": 4.4208,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677255672471097e-05,
+      "loss": 4.4291,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676417077720045e-05,
+      "loss": 4.4206,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675578482968993e-05,
+      "loss": 4.4216,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674739888217941e-05,
+      "loss": 4.4166,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673901293466889e-05,
+      "loss": 4.4095,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673062698715837e-05,
+      "loss": 4.4103,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722257418451576e-05,
+      "loss": 4.4116,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713871470941056e-05,
+      "loss": 4.411,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705485523430536e-05,
+      "loss": 4.4158,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697099575920016e-05,
+      "loss": 4.4126,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688746386016954e-05,
+      "loss": 4.3971,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680360438506434e-05,
+      "loss": 4.4142,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667197449099592e-05,
+      "loss": 4.4195,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66635885434854e-05,
+      "loss": 4.4081,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665520259597488e-05,
+      "loss": 4.4121,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664681664846436e-05,
+      "loss": 4.3982,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663843070095384e-05,
+      "loss": 4.4058,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663004475344332e-05,
+      "loss": 4.4069,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662167518473653e-05,
+      "loss": 4.3852,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661328923722601e-05,
+      "loss": 4.3947,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660491966851922e-05,
+      "loss": 4.4023,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65965337210087e-05,
+      "loss": 4.4029,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658814777349818e-05,
+      "loss": 4.3942,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657976182598766e-05,
+      "loss": 4.4022,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657137587847714e-05,
+      "loss": 4.386,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656298993096662e-05,
+      "loss": 4.4058,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65546039834561e-05,
+      "loss": 4.3896,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546234414749315e-05,
+      "loss": 4.3886,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537848467238795e-05,
+      "loss": 4.4033,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529462519728275e-05,
+      "loss": 4.4008,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521076572217755e-05,
+      "loss": 4.3983,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512690624707234e-05,
+      "loss": 4.3893,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504321056000444e-05,
+      "loss": 4.3921,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.3986,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.3874,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647917959227261e-05,
+      "loss": 4.3909,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647079364476209e-05,
+      "loss": 4.3843,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646240769725157e-05,
+      "loss": 4.3946,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645402174974105e-05,
+      "loss": 4.3822,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644563580223053e-05,
+      "loss": 4.3833,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643724985472001e-05,
+      "loss": 4.3703,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642886390720949e-05,
+      "loss": 4.3903,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642047795969897e-05,
+      "loss": 4.3727,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641209201218845e-05,
+      "loss": 4.3949,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640372244348166e-05,
+      "loss": 4.3916,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639533649597114e-05,
+      "loss": 4.3851,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638695054846062e-05,
+      "loss": 4.3711,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63785646009501e-05,
+      "loss": 4.3901,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637017865343958e-05,
+      "loss": 4.3862,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636180908473279e-05,
+      "loss": 4.3889,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635342313722228e-05,
+      "loss": 4.3756,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634503718971176e-05,
+      "loss": 4.3829,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633665124220124e-05,
+      "loss": 4.3852,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328281673494446e-05,
+      "loss": 4.3861,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319895725983926e-05,
+      "loss": 4.3731,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311509778473406e-05,
+      "loss": 4.3787,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303123830962886e-05,
+      "loss": 4.3765,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294737883452366e-05,
+      "loss": 4.3643,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286351935941846e-05,
+      "loss": 4.3815,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277982367235055e-05,
+      "loss": 4.3819,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269596419724535e-05,
+      "loss": 4.3704,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261210472214015e-05,
+      "loss": 4.3736,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252824524703495e-05,
+      "loss": 4.3632,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.343745231628418,
+      "eval_runtime": 328.2346,
+      "eval_samples_per_second": 1162.556,
+      "eval_steps_per_second": 36.331,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244438577192975e-05,
+      "loss": 4.3582,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623605262968246e-05,
+      "loss": 4.3545,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622766668217194e-05,
+      "loss": 4.3785,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621928073466142e-05,
+      "loss": 4.3649,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621091116595463e-05,
+      "loss": 4.3757,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620252521844411e-05,
+      "loss": 4.3602,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619413927093359e-05,
+      "loss": 4.3665,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61857697022268e-05,
+      "loss": 4.3528,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617738375471628e-05,
+      "loss": 4.3638,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616899780720576e-05,
+      "loss": 4.3638,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616061185969524e-05,
+      "loss": 4.3742,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615222591218472e-05,
+      "loss": 4.3716,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614385634347793e-05,
+      "loss": 4.3559,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135470395967415e-05,
+      "loss": 4.3548,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127084448456895e-05,
+      "loss": 4.3529,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611869850094637e-05,
+      "loss": 4.345,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110328932239584e-05,
+      "loss": 4.3543,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6101942984729064e-05,
+      "loss": 4.3519,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093557037218544e-05,
+      "loss": 4.3496,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608517108970802e-05,
+      "loss": 4.3747,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60767851421975e-05,
+      "loss": 4.3539,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606839919468698e-05,
+      "loss": 4.3519,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606001324717646e-05,
+      "loss": 4.3597,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605162729966594e-05,
+      "loss": 4.3659,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604325773095915e-05,
+      "loss": 4.342,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603487178344863e-05,
+      "loss": 4.346,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602648583593811e-05,
+      "loss": 4.3571,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601809988842759e-05,
+      "loss": 4.354,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60097303197208e-05,
+      "loss": 4.3363,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600134437221028e-05,
+      "loss": 4.3369,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599297480350349e-05,
+      "loss": 4.355,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598458885599297e-05,
+      "loss": 4.3561,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597620290848245e-05,
+      "loss": 4.3498,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596781696097193e-05,
+      "loss": 4.3503,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595944739226514e-05,
+      "loss": 4.3477,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595106144475462e-05,
+      "loss": 4.3481,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59426754972441e-05,
+      "loss": 4.3318,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934305928537316e-05,
+      "loss": 4.3479,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5925919981026796e-05,
+      "loss": 4.3261,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917534033516276e-05,
+      "loss": 4.3407,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909148086005756e-05,
+      "loss": 4.3355,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900762138495235e-05,
+      "loss": 4.3502,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892376190984715e-05,
+      "loss": 4.3379,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883990243474195e-05,
+      "loss": 4.34,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875604295963675e-05,
+      "loss": 4.3393,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867218348453155e-05,
+      "loss": 4.342,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858832400942635e-05,
+      "loss": 4.3474,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850446453432115e-05,
+      "loss": 4.3233,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842060505921595e-05,
+      "loss": 4.3271,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833690937214804e-05,
+      "loss": 4.3436,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825304989704284e-05,
+      "loss": 4.3392,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581691904219377e-05,
+      "loss": 4.3292,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580853309468325e-05,
+      "loss": 4.3238,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580016352597646e-05,
+      "loss": 4.3305,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579177757846594e-05,
+      "loss": 4.3166,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578339163095542e-05,
+      "loss": 4.34,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577502206224863e-05,
+      "loss": 4.3169,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576663611473811e-05,
+      "loss": 4.3267,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575825016722759e-05,
+      "loss": 4.3306,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574986421971707e-05,
+      "loss": 4.3197,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574147827220655e-05,
+      "loss": 4.3219,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573309232469602e-05,
+      "loss": 4.3198,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572470637718551e-05,
+      "loss": 4.312,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716336808478725e-05,
+      "loss": 4.321,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5707950860968205e-05,
+      "loss": 4.3309,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569956491345768e-05,
+      "loss": 4.3162,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569117896594716e-05,
+      "loss": 4.3148,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568279301843664e-05,
+      "loss": 4.3051,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567440707092612e-05,
+      "loss": 4.3141,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56660211234156e-05,
+      "loss": 4.3257,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565763517590508e-05,
+      "loss": 4.3327,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564926560719829e-05,
+      "loss": 4.3141,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640879659687767e-05,
+      "loss": 4.327,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632493712177247e-05,
+      "loss": 4.3327,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562412414347046e-05,
+      "loss": 4.3246,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561573819595994e-05,
+      "loss": 4.3012,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560735224844942e-05,
+      "loss": 4.3215,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55989663009389e-05,
+      "loss": 4.3146,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559058035342838e-05,
+      "loss": 4.3092,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558219440591786e-05,
+      "loss": 4.3135,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557380845840734e-05,
+      "loss": 4.3139,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556543888970055e-05,
+      "loss": 4.324,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555706932099376e-05,
+      "loss": 4.3156,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554868337348324e-05,
+      "loss": 4.3013,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554029742597272e-05,
+      "loss": 4.3117,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55319114784622e-05,
+      "loss": 4.3079,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552352553095168e-05,
+      "loss": 4.3194,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551513958344116e-05,
+      "loss": 4.3177,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550675363593064e-05,
+      "loss": 4.3097,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549836768842013e-05,
+      "loss": 4.3114,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489998119713336e-05,
+      "loss": 4.2999,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481612172202816e-05,
+      "loss": 4.3035,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473226224692296e-05,
+      "loss": 4.3032,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464840277181776e-05,
+      "loss": 4.306,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456470708474985e-05,
+      "loss": 4.3085,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448084760964465e-05,
+      "loss": 4.3157,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439698813453945e-05,
+      "loss": 4.2901,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431312865943425e-05,
+      "loss": 4.3089,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422943297236634e-05,
+      "loss": 4.3164,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414557349726114e-05,
+      "loss": 4.3068,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406171402215594e-05,
+      "loss": 4.3039,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539778545470508e-05,
+      "loss": 4.2956,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538939950719456e-05,
+      "loss": 4.3027,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538102993848777e-05,
+      "loss": 4.3035,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537266036978098e-05,
+      "loss": 4.2835,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536427442227046e-05,
+      "loss": 4.2941,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535590485356367e-05,
+      "loss": 4.2963,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534751890605315e-05,
+      "loss": 4.3036,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533913295854263e-05,
+      "loss": 4.2878,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533074701103211e-05,
+      "loss": 4.3073,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532236106352159e-05,
+      "loss": 4.282,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531397511601107e-05,
+      "loss": 4.3052,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530558916850055e-05,
+      "loss": 4.2927,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297203220990034e-05,
+      "loss": 4.2862,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5288817273479514e-05,
+      "loss": 4.3034,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528043132596899e-05,
+      "loss": 4.3031,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52720617572622e-05,
+      "loss": 4.2978,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526367580975168e-05,
+      "loss": 4.2921,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525530624104489e-05,
+      "loss": 4.2948,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524692029353437e-05,
+      "loss": 4.2978,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523853434602385e-05,
+      "loss": 4.2866,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523014839851333e-05,
+      "loss": 4.292,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221762451002805e-05,
+      "loss": 4.2893,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213376503492285e-05,
+      "loss": 4.2924,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520499055598177e-05,
+      "loss": 4.2888,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519660460847125e-05,
+      "loss": 4.2883,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518821866096073e-05,
+      "loss": 4.2688,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517983271345021e-05,
+      "loss": 4.2931,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517144676593969e-05,
+      "loss": 4.2806,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51630771972329e-05,
+      "loss": 4.2966,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515470762852611e-05,
+      "loss": 4.2939,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514632168101559e-05,
+      "loss": 4.2925,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513793573350507e-05,
+      "loss": 4.2748,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512954978599455e-05,
+      "loss": 4.2949,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512116383848403e-05,
+      "loss": 4.29,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511277789097351e-05,
+      "loss": 4.2965,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510439194346299e-05,
+      "loss": 4.2771,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509600599595247e-05,
+      "loss": 4.2948,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087636427245685e-05,
+      "loss": 4.2924,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079250479735165e-05,
+      "loss": 4.2949,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070864532224645e-05,
+      "loss": 4.2747,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062478584714125e-05,
+      "loss": 4.2885,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054109016007334e-05,
+      "loss": 4.2848,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045723068496814e-05,
+      "loss": 4.276,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037337120986294e-05,
+      "loss": 4.2854,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028951173475774e-05,
+      "loss": 4.2895,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020581604768983e-05,
+      "loss": 4.2807,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501219565725846e-05,
+      "loss": 4.2798,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500382608855168e-05,
+      "loss": 4.278,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.260278701782227,
+      "eval_runtime": 328.791,
+      "eval_samples_per_second": 1160.588,
+      "eval_steps_per_second": 36.269,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499544014104116e-05,
+      "loss": 4.2622,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498705419353064e-05,
+      "loss": 4.2658,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497866824602012e-05,
+      "loss": 4.2855,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49702822985096e-05,
+      "loss": 4.2755,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496189635099908e-05,
+      "loss": 4.286,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495351040348856e-05,
+      "loss": 4.2729,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494512445597804e-05,
+      "loss": 4.2768,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493673850846752e-05,
+      "loss": 4.2632,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492835256095699e-05,
+      "loss": 4.2761,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491996661344647e-05,
+      "loss": 4.2754,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491158066593595e-05,
+      "loss": 4.2799,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490319471842543e-05,
+      "loss": 4.286,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489482514971865e-05,
+      "loss": 4.2669,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488643920220813e-05,
+      "loss": 4.2688,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487805325469761e-05,
+      "loss": 4.2649,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486966730718709e-05,
+      "loss": 4.2571,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486128135967657e-05,
+      "loss": 4.2686,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485289541216605e-05,
+      "loss": 4.2652,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484450946465553e-05,
+      "loss": 4.2648,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483612351714501e-05,
+      "loss": 4.2857,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482773756963449e-05,
+      "loss": 4.2688,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481935162212397e-05,
+      "loss": 4.2651,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481096567461345e-05,
+      "loss": 4.2752,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480257972710293e-05,
+      "loss": 4.2793,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4794210158396136e-05,
+      "loss": 4.2584,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4785840589689346e-05,
+      "loss": 4.2649,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777454642178825e-05,
+      "loss": 4.269,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476906869466831e-05,
+      "loss": 4.267,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476068274715779e-05,
+      "loss": 4.2512,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475229679964727e-05,
+      "loss": 4.2547,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474392723094048e-05,
+      "loss": 4.2696,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473554128342996e-05,
+      "loss": 4.271,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472715533591944e-05,
+      "loss": 4.2668,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471876938840892e-05,
+      "loss": 4.2705,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47103834408984e-05,
+      "loss": 4.2629,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470199749338788e-05,
+      "loss": 4.2677,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693611545877354e-05,
+      "loss": 4.2515,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685225598366834e-05,
+      "loss": 4.2605,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467685602966005e-05,
+      "loss": 4.243,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466847008214953e-05,
+      "loss": 4.2558,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466008413463901e-05,
+      "loss": 4.2533,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465169818712849e-05,
+      "loss": 4.2679,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643328618421706e-05,
+      "loss": 4.2569,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634959049714915e-05,
+      "loss": 4.2582,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626573102204395e-05,
+      "loss": 4.2573,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618187154693875e-05,
+      "loss": 4.2604,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609801207183355e-05,
+      "loss": 4.2653,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601431638476564e-05,
+      "loss": 4.2443,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593045690966044e-05,
+      "loss": 4.2445,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584659743455524e-05,
+      "loss": 4.2633,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576273795945004e-05,
+      "loss": 4.2593,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567887848434484e-05,
+      "loss": 4.2475,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559501900923964e-05,
+      "loss": 4.2438,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551115953413444e-05,
+      "loss": 4.2542,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454274638470665e-05,
+      "loss": 4.2379,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453437681599987e-05,
+      "loss": 4.2586,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452599086848935e-05,
+      "loss": 4.239,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451760492097883e-05,
+      "loss": 4.2479,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45092189734683e-05,
+      "loss": 4.2494,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450083302595778e-05,
+      "loss": 4.2441,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449244707844726e-05,
+      "loss": 4.2401,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448406113093674e-05,
+      "loss": 4.2442,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447567518342622e-05,
+      "loss": 4.234,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446730561471944e-05,
+      "loss": 4.2436,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445891966720892e-05,
+      "loss": 4.2526,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44505337196984e-05,
+      "loss": 4.2388,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444214777218788e-05,
+      "loss": 4.237,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443376182467736e-05,
+      "loss": 4.228,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442537587716684e-05,
+      "loss": 4.2338,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4417006308460046e-05,
+      "loss": 4.2506,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408620360949526e-05,
+      "loss": 4.2517,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400234413439006e-05,
+      "loss": 4.2412,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391848465928486e-05,
+      "loss": 4.249,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383462518417966e-05,
+      "loss": 4.2587,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4375076570907446e-05,
+      "loss": 4.2461,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4366707002200655e-05,
+      "loss": 4.228,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435832105469014e-05,
+      "loss": 4.2434,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434993510717962e-05,
+      "loss": 4.2408,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43415491596691e-05,
+      "loss": 4.2334,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433316321215858e-05,
+      "loss": 4.2371,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432477726464806e-05,
+      "loss": 4.243,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431639131713754e-05,
+      "loss": 4.25,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308005369627015e-05,
+      "loss": 4.2408,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429963580092023e-05,
+      "loss": 4.2219,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429124985340971e-05,
+      "loss": 4.2389,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428286390589919e-05,
+      "loss": 4.236,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274477958388664e-05,
+      "loss": 4.2397,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426610838968188e-05,
+      "loss": 4.2414,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425772244217136e-05,
+      "loss": 4.238,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249352873464575e-05,
+      "loss": 4.2381,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240966925954055e-05,
+      "loss": 4.2288,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232580978443535e-05,
+      "loss": 4.2257,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224195030933015e-05,
+      "loss": 4.2297,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421580908342249e-05,
+      "loss": 4.2379,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420742313591197e-05,
+      "loss": 4.2337,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419903718840145e-05,
+      "loss": 4.2401,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419065124089093e-05,
+      "loss": 4.2185,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418228167218414e-05,
+      "loss": 4.2328,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417389572467362e-05,
+      "loss": 4.242,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41655097771631e-05,
+      "loss": 4.2364,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415712382965258e-05,
+      "loss": 4.2319,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414873788214206e-05,
+      "loss": 4.2258,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414036831343527e-05,
+      "loss": 4.2283,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413198236592475e-05,
+      "loss": 4.2336,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412359641841423e-05,
+      "loss": 4.2107,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411521047090371e-05,
+      "loss": 4.2266,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410684090219692e-05,
+      "loss": 4.223,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409847133349014e-05,
+      "loss": 4.2313,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409008538597961e-05,
+      "loss": 4.2159,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408169943846909e-05,
+      "loss": 4.2352,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407331349095857e-05,
+      "loss": 4.2213,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406492754344805e-05,
+      "loss": 4.2301,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405654159593753e-05,
+      "loss": 4.2231,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404815564842701e-05,
+      "loss": 4.214,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40397697009165e-05,
+      "loss": 4.2327,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403140013220971e-05,
+      "loss": 4.2366,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402301418469919e-05,
+      "loss": 4.23,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014644615992396e-05,
+      "loss": 4.2214,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4006258668481876e-05,
+      "loss": 4.2241,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3997872720971356e-05,
+      "loss": 4.2297,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3989486773460836e-05,
+      "loss": 4.2165,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981100825950316e-05,
+      "loss": 4.2237,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972714878439796e-05,
+      "loss": 4.2175,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964328930929276e-05,
+      "loss": 4.2213,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3955942983418755e-05,
+      "loss": 4.2221,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947573414711965e-05,
+      "loss": 4.2224,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393918746720145e-05,
+      "loss": 4.204,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393081789849466e-05,
+      "loss": 4.2218,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392243195098414e-05,
+      "loss": 4.2141,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391404600347362e-05,
+      "loss": 4.2315,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39056600559631e-05,
+      "loss": 4.2217,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389727410845258e-05,
+      "loss": 4.227,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388888816094206e-05,
+      "loss": 4.2072,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388050221343154e-05,
+      "loss": 4.2287,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387211626592102e-05,
+      "loss": 4.2257,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38637303184105e-05,
+      "loss": 4.2255,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385536074970371e-05,
+      "loss": 4.2124,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384697480219319e-05,
+      "loss": 4.2282,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383858885468267e-05,
+      "loss": 4.2252,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383020290717215e-05,
+      "loss": 4.2321,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821833338465365e-05,
+      "loss": 4.2068,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813447390954845e-05,
+      "loss": 4.2212,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805061443444325e-05,
+      "loss": 4.2178,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37966754959338e-05,
+      "loss": 4.2073,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788305927227014e-05,
+      "loss": 4.2205,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779919979716494e-05,
+      "loss": 4.2209,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771534032205974e-05,
+      "loss": 4.2185,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376314808469545e-05,
+      "loss": 4.2124,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375477851598866e-05,
+      "loss": 4.2149,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.203587055206299,
+      "eval_runtime": 296.7852,
+      "eval_samples_per_second": 1285.748,
+      "eval_steps_per_second": 40.181,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374639256847814e-05,
+      "loss": 4.2043,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373800662096762e-05,
+      "loss": 4.198,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37296206734571e-05,
+      "loss": 4.2213,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372123472594658e-05,
+      "loss": 4.2128,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371284877843606e-05,
+      "loss": 4.2188,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370446283092554e-05,
+      "loss": 4.2127,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369607688341502e-05,
+      "loss": 4.2088,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368770731470823e-05,
+      "loss": 4.2029,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367932136719771e-05,
+      "loss": 4.2103,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367093541968719e-05,
+      "loss": 4.2122,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366254947217667e-05,
+      "loss": 4.215,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365417990346988e-05,
+      "loss": 4.2195,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645810334763096e-05,
+      "loss": 4.2031,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637424387252576e-05,
+      "loss": 4.2069,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3629038439742056e-05,
+      "loss": 4.2051,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620652492231536e-05,
+      "loss": 4.1917,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612266544721016e-05,
+      "loss": 4.2074,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3603880597210496e-05,
+      "loss": 4.2046,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595494649699976e-05,
+      "loss": 4.1992,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3587108702189456e-05,
+      "loss": 4.2193,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3578739133482665e-05,
+      "loss": 4.2105,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3570353185972145e-05,
+      "loss": 4.2045,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3561967238461625e-05,
+      "loss": 4.211,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3553581290951105e-05,
+      "loss": 4.2192,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545195343440585e-05,
+      "loss": 4.1942,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3536825774733794e-05,
+      "loss": 4.2036,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528439827223274e-05,
+      "loss": 4.2103,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352005387971276e-05,
+      "loss": 4.1991,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351166793220224e-05,
+      "loss": 4.1913,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350329836349545e-05,
+      "loss": 4.1946,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349491241598493e-05,
+      "loss": 4.2098,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348652646847441e-05,
+      "loss": 4.2103,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347814052096389e-05,
+      "loss": 4.2077,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34697709522571e-05,
+      "loss": 4.21,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346138500474658e-05,
+      "loss": 4.2027,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345299905723606e-05,
+      "loss": 4.2023,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344461310972554e-05,
+      "loss": 4.1954,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343624354101875e-05,
+      "loss": 4.1987,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342785759350823e-05,
+      "loss": 4.1848,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419488024801444e-05,
+      "loss": 4.1933,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411102077290924e-05,
+      "loss": 4.1957,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402716129780404e-05,
+      "loss": 4.2079,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394330182269883e-05,
+      "loss": 4.1987,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385944234759363e-05,
+      "loss": 4.1998,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377558287248843e-05,
+      "loss": 4.2012,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336917233973832e-05,
+      "loss": 4.1963,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336080277103153e-05,
+      "loss": 4.2071,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335243320232474e-05,
+      "loss": 4.1901,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334404725481422e-05,
+      "loss": 4.1844,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33356613073037e-05,
+      "loss": 4.2038,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332727535979318e-05,
+      "loss": 4.2005,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331888941228267e-05,
+      "loss": 4.1893,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331050346477215e-05,
+      "loss": 4.1902,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330211751726163e-05,
+      "loss": 4.1958,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329373156975111e-05,
+      "loss": 4.1781,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328536200104432e-05,
+      "loss": 4.1998,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32769760535338e-05,
+      "loss": 4.1821,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326859010602328e-05,
+      "loss": 4.1858,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326020415851275e-05,
+      "loss": 4.1953,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325181821100223e-05,
+      "loss": 4.1865,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324343226349171e-05,
+      "loss": 4.1843,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3235062694784926e-05,
+      "loss": 4.1866,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3226676747274406e-05,
+      "loss": 4.1769,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218290799763886e-05,
+      "loss": 4.1833,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209904852253366e-05,
+      "loss": 4.1935,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201518904742846e-05,
+      "loss": 4.1844,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193132957232326e-05,
+      "loss": 4.1797,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184747009721806e-05,
+      "loss": 4.1755,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176361062211286e-05,
+      "loss": 4.1744,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167991493504495e-05,
+      "loss": 4.1903,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159605545993975e-05,
+      "loss": 4.1943,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151219598483455e-05,
+      "loss": 4.1878,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3142833650972935e-05,
+      "loss": 4.1917,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3134464082266144e-05,
+      "loss": 4.1984,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3126078134755624e-05,
+      "loss": 4.1976,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3117692187245104e-05,
+      "loss": 4.1711,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310932261853832e-05,
+      "loss": 4.1829,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31009366710278e-05,
+      "loss": 4.1896,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309255072351728e-05,
+      "loss": 4.1729,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308416477600676e-05,
+      "loss": 4.1873,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307577882849624e-05,
+      "loss": 4.1808,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306739288098572e-05,
+      "loss": 4.1948,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30590069334752e-05,
+      "loss": 4.1878,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305063736476841e-05,
+      "loss": 4.1667,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304225141725789e-05,
+      "loss": 4.1838,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303386546974737e-05,
+      "loss": 4.1807,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302547952223685e-05,
+      "loss": 4.1885,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301709357472633e-05,
+      "loss": 4.1825,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300870762721581e-05,
+      "loss": 4.1868,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300032167970529e-05,
+      "loss": 4.1829,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299193573219477e-05,
+      "loss": 4.1742,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983566163487984e-05,
+      "loss": 4.1705,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975180215977464e-05,
+      "loss": 4.1761,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296681064727067e-05,
+      "loss": 4.1836,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295842469976015e-05,
+      "loss": 4.1801,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295003875224963e-05,
+      "loss": 4.1828,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294165280473911e-05,
+      "loss": 4.1679,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933266857228586e-05,
+      "loss": 4.1788,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924880909718066e-05,
+      "loss": 4.1861,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916494962207546e-05,
+      "loss": 4.1852,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908109014697026e-05,
+      "loss": 4.1784,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289973944599024e-05,
+      "loss": 4.1752,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289135349847972e-05,
+      "loss": 4.1722,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28829675509692e-05,
+      "loss": 4.1834,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287458160345868e-05,
+      "loss": 4.1538,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286621203475189e-05,
+      "loss": 4.1734,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285782608724137e-05,
+      "loss": 4.1664,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284944013973085e-05,
+      "loss": 4.1802,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284105419222033e-05,
+      "loss": 4.163,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283266824470981e-05,
+      "loss": 4.1874,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282428229719929e-05,
+      "loss": 4.1686,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281589634968877e-05,
+      "loss": 4.1785,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280752678098198e-05,
+      "loss": 4.1694,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279914083347146e-05,
+      "loss": 4.1655,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790754885960946e-05,
+      "loss": 4.1766,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782368938450426e-05,
+      "loss": 4.1849,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2773999369743635e-05,
+      "loss": 4.1771,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765613422233115e-05,
+      "loss": 4.1676,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757227474722595e-05,
+      "loss": 4.1709,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748841527212075e-05,
+      "loss": 4.1792,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740455579701555e-05,
+      "loss": 4.1648,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2732069632191035e-05,
+      "loss": 4.1754,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2723683684680515e-05,
+      "loss": 4.1624,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715297737169995e-05,
+      "loss": 4.1696,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2706928168463204e-05,
+      "loss": 4.1707,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698542220952684e-05,
+      "loss": 4.1726,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690156273442164e-05,
+      "loss": 4.1507,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2681770325931644e-05,
+      "loss": 4.1719,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267338437842113e-05,
+      "loss": 4.1625,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266501480971434e-05,
+      "loss": 4.1781,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265664524100755e-05,
+      "loss": 4.1722,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264825929349703e-05,
+      "loss": 4.1765,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263987334598651e-05,
+      "loss": 4.1555,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263148739847599e-05,
+      "loss": 4.1777,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262310145096547e-05,
+      "loss": 4.1747,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261471550345495e-05,
+      "loss": 4.1729,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260632955594442e-05,
+      "loss": 4.1657,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25979436084339e-05,
+      "loss": 4.173,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258957403972712e-05,
+      "loss": 4.1759,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25811880922166e-05,
+      "loss": 4.1797,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257280214470608e-05,
+      "loss": 4.1586,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256441619719556e-05,
+      "loss": 4.1708,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255604662848877e-05,
+      "loss": 4.1679,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547660680978247e-05,
+      "loss": 4.157,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539274733467727e-05,
+      "loss": 4.1736,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2530888785957206e-05,
+      "loss": 4.1674,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252251921725042e-05,
+      "loss": 4.1764,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514133269739896e-05,
+      "loss": 4.1605,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2505747322229375e-05,
+      "loss": 4.1648,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.162723064422607,
+      "eval_runtime": 299.403,
+      "eval_samples_per_second": 1274.506,
+      "eval_steps_per_second": 39.829,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497361374718855e-05,
+      "loss": 4.1537,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488975427208335e-05,
+      "loss": 4.1466,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248058947969782e-05,
+      "loss": 4.1712,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24722035321873e-05,
+      "loss": 4.1649,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246381758467678e-05,
+      "loss": 4.1697,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245543163716626e-05,
+      "loss": 4.1636,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244704568965574e-05,
+      "loss": 4.1542,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243865974214522e-05,
+      "loss": 4.1579,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24302737946347e-05,
+      "loss": 4.1597,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242188784712418e-05,
+      "loss": 4.1612,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241350189961366e-05,
+      "loss": 4.1696,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405115952103135e-05,
+      "loss": 4.1723,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239674638339635e-05,
+      "loss": 4.1567,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238836043588583e-05,
+      "loss": 4.1572,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237997448837531e-05,
+      "loss": 4.1597,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371588540864784e-05,
+      "loss": 4.1425,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236320259335427e-05,
+      "loss": 4.1595,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235481664584375e-05,
+      "loss": 4.1544,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234643069833323e-05,
+      "loss": 4.1532,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233804475082271e-05,
+      "loss": 4.1694,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232965880331219e-05,
+      "loss": 4.1633,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232127285580167e-05,
+      "loss": 4.1602,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231288690829115e-05,
+      "loss": 4.1607,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230450096078063e-05,
+      "loss": 4.1699,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229611501327011e-05,
+      "loss": 4.148,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228774544456332e-05,
+      "loss": 4.1601,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22793594970528e-05,
+      "loss": 4.1564,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227097354954228e-05,
+      "loss": 4.1514,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226258760203176e-05,
+      "loss": 4.1496,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225421803332497e-05,
+      "loss": 4.143,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245848464618184e-05,
+      "loss": 4.1643,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237462517107664e-05,
+      "loss": 4.1617,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229076569597144e-05,
+      "loss": 4.1583,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220690622086624e-05,
+      "loss": 4.1638,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221232105337983e-05,
+      "loss": 4.1532,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220393510586931e-05,
+      "loss": 4.1575,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219554915835879e-05,
+      "loss": 4.1525,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218716321084827e-05,
+      "loss": 4.1479,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217877726333775e-05,
+      "loss": 4.1391,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217039131582723e-05,
+      "loss": 4.1457,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216202174712044e-05,
+      "loss": 4.1508,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215363579960992e-05,
+      "loss": 4.1627,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214524985209941e-05,
+      "loss": 4.1505,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213686390458889e-05,
+      "loss": 4.1507,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212847795707837e-05,
+      "loss": 4.1595,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212009200956785e-05,
+      "loss": 4.1506,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211172244086106e-05,
+      "loss": 4.1615,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210333649335054e-05,
+      "loss": 4.1476,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209495054584002e-05,
+      "loss": 4.1351,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20865645983295e-05,
+      "loss": 4.1576,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207817865081897e-05,
+      "loss": 4.1523,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206980908211219e-05,
+      "loss": 4.1496,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061423134601667e-05,
+      "loss": 4.1396,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053037187091147e-05,
+      "loss": 4.1506,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044651239580626e-05,
+      "loss": 4.1294,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036265292070106e-05,
+      "loss": 4.1534,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027879344559586e-05,
+      "loss": 4.1387,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019493397049066e-05,
+      "loss": 4.1385,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011107449538546e-05,
+      "loss": 4.154,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2002721502028026e-05,
+      "loss": 4.1381,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994351933321235e-05,
+      "loss": 4.1396,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1985965985810715e-05,
+      "loss": 4.1448,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1977580038300195e-05,
+      "loss": 4.1304,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196922684839714e-05,
+      "loss": 4.1415,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196084090088662e-05,
+      "loss": 4.1463,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19524549533761e-05,
+      "loss": 4.142,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194406900586558e-05,
+      "loss": 4.1358,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193568305835506e-05,
+      "loss": 4.1314,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192729711084454e-05,
+      "loss": 4.1289,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191891116333402e-05,
+      "loss": 4.1423,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19105252158235e-05,
+      "loss": 4.1504,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190213926831298e-05,
+      "loss": 4.1411,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189375332080246e-05,
+      "loss": 4.1489,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188536737329194e-05,
+      "loss": 4.1511,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187698142578142e-05,
+      "loss": 4.1545,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186862823587836e-05,
+      "loss": 4.1288,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186024228836784e-05,
+      "loss": 4.143,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185185634085732e-05,
+      "loss": 4.1441,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18434703933468e-05,
+      "loss": 4.1233,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183508444583628e-05,
+      "loss": 4.146,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826698498325765e-05,
+      "loss": 4.1396,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818312550815245e-05,
+      "loss": 4.146,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809926603304725e-05,
+      "loss": 4.1437,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801557034597934e-05,
+      "loss": 4.1255,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793171087087414e-05,
+      "loss": 4.1372,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784785139576894e-05,
+      "loss": 4.1375,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776399192066373e-05,
+      "loss": 4.1417,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176802962335958e-05,
+      "loss": 4.1388,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175967643345652e-05,
+      "loss": 4.1447,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751290485946e-05,
+      "loss": 4.137,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174290453843549e-05,
+      "loss": 4.1338,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173451859092497e-05,
+      "loss": 4.1278,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172613264341445e-05,
+      "loss": 4.1301,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171774669590393e-05,
+      "loss": 4.1454,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170936074839341e-05,
+      "loss": 4.139,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170097480088289e-05,
+      "loss": 4.1373,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169258885337237e-05,
+      "loss": 4.1268,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168420290586185e-05,
+      "loss": 4.134,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167581695835133e-05,
+      "loss": 4.1425,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166743101084081e-05,
+      "loss": 4.14,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165904506333028e-05,
+      "loss": 4.1372,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1650675494623496e-05,
+      "loss": 4.1322,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642289547112976e-05,
+      "loss": 4.1323,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633903599602456e-05,
+      "loss": 4.1386,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625517652091936e-05,
+      "loss": 4.1142,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161716446218888e-05,
+      "loss": 4.128,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160877851467836e-05,
+      "loss": 4.1238,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160039256716784e-05,
+      "loss": 4.1401,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159200661965732e-05,
+      "loss": 4.126,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158363705095053e-05,
+      "loss": 4.142,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157525110344001e-05,
+      "loss": 4.1278,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156686515592949e-05,
+      "loss": 4.1329,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155847920841897e-05,
+      "loss": 4.1335,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155010963971218e-05,
+      "loss": 4.1187,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154172369220166e-05,
+      "loss": 4.1341,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153333774469114e-05,
+      "loss": 4.1431,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524951797180626e-05,
+      "loss": 4.1346,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516565849670106e-05,
+      "loss": 4.134,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150817990215958e-05,
+      "loss": 4.1214,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149979395464906e-05,
+      "loss": 4.1375,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149140800713854e-05,
+      "loss": 4.1295,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1483038438431755e-05,
+      "loss": 4.1309,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474668869724964e-05,
+      "loss": 4.1192,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466282922214444e-05,
+      "loss": 4.1281,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457896974703924e-05,
+      "loss": 4.1329,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14495110271934e-05,
+      "loss": 4.1311,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144114145848661e-05,
+      "loss": 4.1095,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143275551097609e-05,
+      "loss": 4.1347,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142436956346558e-05,
+      "loss": 4.1175,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141598361595505e-05,
+      "loss": 4.1389,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140759766844453e-05,
+      "loss": 4.1284,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139922809973775e-05,
+      "loss": 4.1363,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139084215222723e-05,
+      "loss": 4.1155,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13824562047167e-05,
+      "loss": 4.1401,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137407025720618e-05,
+      "loss": 4.1329,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136568430969566e-05,
+      "loss": 4.1323,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135729836218514e-05,
+      "loss": 4.121,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134892879347835e-05,
+      "loss": 4.1331,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134054284596783e-05,
+      "loss": 4.1346,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133215689845732e-05,
+      "loss": 4.1378,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13237709509468e-05,
+      "loss": 4.1203,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131538500343628e-05,
+      "loss": 4.1297,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130699905592576e-05,
+      "loss": 4.1306,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129861310841524e-05,
+      "loss": 4.1225,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129022716090472e-05,
+      "loss": 4.1268,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12818412133942e-05,
+      "loss": 4.127,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273471644687406e-05,
+      "loss": 4.1331,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1265085697176886e-05,
+      "loss": 4.1238,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1256699749666366e-05,
+      "loss": 4.1243,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.129836559295654,
+      "eval_runtime": 582.9613,
+      "eval_samples_per_second": 654.573,
+      "eval_steps_per_second": 20.456,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1248313802155846e-05,
+      "loss": 4.1177,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1239927854645326e-05,
+      "loss": 4.1059,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1231558285938535e-05,
+      "loss": 4.1285,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1223172338428015e-05,
+      "loss": 4.127,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1214786390917495e-05,
+      "loss": 4.126,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120640044340698e-05,
+      "loss": 4.1259,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119804725350392e-05,
+      "loss": 4.1168,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11896613059934e-05,
+      "loss": 4.1157,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118127535848288e-05,
+      "loss": 4.1212,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117288941097236e-05,
+      "loss": 4.1205,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116450346346184e-05,
+      "loss": 4.1292,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115611751595132e-05,
+      "loss": 4.1316,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114774794724453e-05,
+      "loss": 4.1182,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113936199973401e-05,
+      "loss": 4.1186,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113097605222349e-05,
+      "loss": 4.1174,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112259010471297e-05,
+      "loss": 4.1015,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111420415720245e-05,
+      "loss": 4.1208,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105818209691935e-05,
+      "loss": 4.116,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1097432262181415e-05,
+      "loss": 4.1179,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108904631467089e-05,
+      "loss": 4.1254,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108066036716037e-05,
+      "loss": 4.1289,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107227441964985e-05,
+      "loss": 4.1199,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106388847213933e-05,
+      "loss": 4.1209,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105551890343254e-05,
+      "loss": 4.1296,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104713295592202e-05,
+      "loss": 4.1122,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10387470084115e-05,
+      "loss": 4.1211,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103036106090098e-05,
+      "loss": 4.1196,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102197511339046e-05,
+      "loss": 4.1148,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101360554468367e-05,
+      "loss": 4.1094,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100523597597689e-05,
+      "loss": 4.1092,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099685002846636e-05,
+      "loss": 4.1209,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098848045975958e-05,
+      "loss": 4.1267,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098009451224906e-05,
+      "loss": 4.1202,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097170856473853e-05,
+      "loss": 4.1241,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096332261722801e-05,
+      "loss": 4.1118,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095493666971749e-05,
+      "loss": 4.1205,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094655072220697e-05,
+      "loss": 4.1207,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093816477469645e-05,
+      "loss": 4.1028,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092977882718593e-05,
+      "loss": 4.1064,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092139287967541e-05,
+      "loss": 4.1058,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091300693216489e-05,
+      "loss": 4.1103,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090462098465437e-05,
+      "loss": 4.1257,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089623503714386e-05,
+      "loss": 4.1125,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0887881847240796e-05,
+      "loss": 4.1097,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879495899730276e-05,
+      "loss": 4.1247,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0871109952219755e-05,
+      "loss": 4.1145,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0862724004709235e-05,
+      "loss": 4.1212,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854354436002445e-05,
+      "loss": 4.1094,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0845968488491925e-05,
+      "loss": 4.0974,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837582540981404e-05,
+      "loss": 4.1257,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829196593470884e-05,
+      "loss": 4.1131,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0820810645960364e-05,
+      "loss": 4.1142,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081244107725358e-05,
+      "loss": 4.1083,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080405512974306e-05,
+      "loss": 4.1062,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079566918223254e-05,
+      "loss": 4.0997,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078729961352575e-05,
+      "loss": 4.1094,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077891366601523e-05,
+      "loss": 4.1058,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077052771850471e-05,
+      "loss": 4.1062,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076214177099419e-05,
+      "loss": 4.1148,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075375582348367e-05,
+      "loss": 4.1042,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074536987597315e-05,
+      "loss": 4.1004,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073698392846263e-05,
+      "loss": 4.1132,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072859798095211e-05,
+      "loss": 4.0927,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072021203344159e-05,
+      "loss": 4.1053,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071182608593107e-05,
+      "loss": 4.1099,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070344013842055e-05,
+      "loss": 4.1072,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069505419091003e-05,
+      "loss": 4.0973,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068666824339951e-05,
+      "loss": 4.0974,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678298674692725e-05,
+      "loss": 4.0892,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06699127271822e-05,
+      "loss": 4.1094,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066152677967168e-05,
+      "loss": 4.1153,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653157210964894e-05,
+      "loss": 4.11,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064477126345437e-05,
+      "loss": 4.1072,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063638531594385e-05,
+      "loss": 4.1174,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062799936843333e-05,
+      "loss": 4.1202,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061962979972654e-05,
+      "loss": 4.0902,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611243852216016e-05,
+      "loss": 4.1097,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06028579047055e-05,
+      "loss": 4.1103,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059447195719498e-05,
+      "loss": 4.0853,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058608600968446e-05,
+      "loss": 4.1137,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057770006217394e-05,
+      "loss": 4.1006,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056931411466342e-05,
+      "loss": 4.1143,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05609281671529e-05,
+      "loss": 4.1105,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055255859844611e-05,
+      "loss": 4.0907,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054417265093559e-05,
+      "loss": 4.0983,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053578670342507e-05,
+      "loss": 4.1073,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052740075591455e-05,
+      "loss": 4.1064,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051903118720776e-05,
+      "loss": 4.1027,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051066161850097e-05,
+      "loss": 4.1095,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050227567099045e-05,
+      "loss": 4.1015,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0493889723479936e-05,
+      "loss": 4.0968,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0485503775969416e-05,
+      "loss": 4.0987,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477117828458896e-05,
+      "loss": 4.095,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0468731880948376e-05,
+      "loss": 4.1086,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460345933437856e-05,
+      "loss": 4.1042,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0451959985927336e-05,
+      "loss": 4.1093,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443574038416816e-05,
+      "loss": 4.0877,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435220848513754e-05,
+      "loss": 4.1017,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0426834901003234e-05,
+      "loss": 4.1098,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418448953492714e-05,
+      "loss": 4.1077,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041007938478592e-05,
+      "loss": 4.1018,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04016934372754e-05,
+      "loss": 4.0954,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039330748976489e-05,
+      "loss": 4.1014,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038492154225437e-05,
+      "loss": 4.1041,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037653559474385e-05,
+      "loss": 4.0801,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036814964723333e-05,
+      "loss": 4.0996,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035976369972281e-05,
+      "loss": 4.0872,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035137775221229e-05,
+      "loss": 4.1056,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034299180470177e-05,
+      "loss": 4.094,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033460585719125e-05,
+      "loss": 4.1034,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032621990968073e-05,
+      "loss": 4.0957,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031785034097394e-05,
+      "loss": 4.104,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030946439346342e-05,
+      "loss": 4.0972,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030109482475663e-05,
+      "loss": 4.0849,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029270887724611e-05,
+      "loss": 4.1029,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028432292973559e-05,
+      "loss": 4.1075,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0275936982225074e-05,
+      "loss": 4.1033,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267551034714554e-05,
+      "loss": 4.1032,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025916508720403e-05,
+      "loss": 4.0875,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025077913969351e-05,
+      "loss": 4.1027,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024239319218299e-05,
+      "loss": 4.102,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023400724467247e-05,
+      "loss": 4.093,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022562129716195e-05,
+      "loss": 4.0913,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021723534965143e-05,
+      "loss": 4.0956,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020884940214091e-05,
+      "loss": 4.1001,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200479833434116e-05,
+      "loss": 4.0987,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0192093885923596e-05,
+      "loss": 4.0772,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0183707938413076e-05,
+      "loss": 4.1029,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0175321990902556e-05,
+      "loss": 4.0835,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016693604339204e-05,
+      "loss": 4.1067,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015856647468525e-05,
+      "loss": 4.0994,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015018052717473e-05,
+      "loss": 4.1013,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014179457966421e-05,
+      "loss": 4.0831,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013340863215369e-05,
+      "loss": 4.108,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012502268464317e-05,
+      "loss": 4.1007,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011665311593638e-05,
+      "loss": 4.0993,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010826716842586e-05,
+      "loss": 4.0953,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009988122091534e-05,
+      "loss": 4.0945,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009149527340482e-05,
+      "loss": 4.1031,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00831093258943e-05,
+      "loss": 4.1104,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007473975718751e-05,
+      "loss": 4.0879,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006635380967699e-05,
+      "loss": 4.0969,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0057967862166476e-05,
+      "loss": 4.1027,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0049581914655956e-05,
+      "loss": 4.0898,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041212345949165e-05,
+      "loss": 4.0912,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0032826398438645e-05,
+      "loss": 4.098,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024456829731854e-05,
+      "loss": 4.1025,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016070882221334e-05,
+      "loss": 4.0929,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0007684934710814e-05,
+      "loss": 4.095,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.1065754890441895,
+      "eval_runtime": 604.3498,
+      "eval_samples_per_second": 631.407,
+      "eval_steps_per_second": 19.732,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999298987200294e-05,
+      "loss": 4.0785,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9990913039689774e-05,
+      "loss": 4.0799,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9982527092179254e-05,
+      "loss": 4.0947,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9974141144668734e-05,
+      "loss": 4.0939,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9965755197158214e-05,
+      "loss": 4.1018,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957369249647694e-05,
+      "loss": 4.0906,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9948983302137174e-05,
+      "loss": 4.0882,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9940597354626654e-05,
+      "loss": 4.088,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9932211407116134e-05,
+      "loss": 4.0864,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9923825459605614e-05,
+      "loss": 4.092,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9915439512095094e-05,
+      "loss": 4.0972,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9907053564584574e-05,
+      "loss": 4.0983,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989868399587778e-05,
+      "loss": 4.0877,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989029804836726e-05,
+      "loss": 4.0899,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988191210085674e-05,
+      "loss": 4.0901,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987352615334622e-05,
+      "loss": 4.071,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98651402058357e-05,
+      "loss": 4.0869,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985675425832518e-05,
+      "loss": 4.0855,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984836831081466e-05,
+      "loss": 4.09,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983998236330414e-05,
+      "loss": 4.0934,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983159641579363e-05,
+      "loss": 4.0997,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98232104682831e-05,
+      "loss": 4.0905,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981482452077258e-05,
+      "loss": 4.0925,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98064549520658e-05,
+      "loss": 4.0958,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979806900455528e-05,
+      "loss": 4.085,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978968305704475e-05,
+      "loss": 4.0892,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978129710953423e-05,
+      "loss": 4.0885,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977291116202371e-05,
+      "loss": 4.0846,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976452521451319e-05,
+      "loss": 4.0832,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97561556458064e-05,
+      "loss": 4.0758,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974776969829588e-05,
+      "loss": 4.0882,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973938375078537e-05,
+      "loss": 4.0978,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973099780327485e-05,
+      "loss": 4.0911,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972261185576433e-05,
+      "loss": 4.0899,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9714242287057536e-05,
+      "loss": 4.0825,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9705856339547016e-05,
+      "loss": 4.0944,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697470392036496e-05,
+      "loss": 4.0887,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689084444525976e-05,
+      "loss": 4.0758,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9680698497015456e-05,
+      "loss": 4.0775,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672328928308665e-05,
+      "loss": 4.075,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9663942980798145e-05,
+      "loss": 4.0778,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9655557033287625e-05,
+      "loss": 4.0981,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9647171085777105e-05,
+      "loss": 4.0835,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9638785138266585e-05,
+      "loss": 4.0826,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96304155695598e-05,
+      "loss": 4.0936,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962202962204928e-05,
+      "loss": 4.083,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961364367453876e-05,
+      "loss": 4.0933,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960525772702824e-05,
+      "loss": 4.0836,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959687177951772e-05,
+      "loss": 4.06,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958850221081093e-05,
+      "loss": 4.1003,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958011626330041e-05,
+      "loss": 4.0844,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957173031578989e-05,
+      "loss": 4.0852,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956334436827937e-05,
+      "loss": 4.0768,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955495842076885e-05,
+      "loss": 4.0741,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954657247325833e-05,
+      "loss": 4.0771,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953820290455154e-05,
+      "loss": 4.0789,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952981695704102e-05,
+      "loss": 4.0759,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95214310095305e-05,
+      "loss": 4.0743,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9513045062019985e-05,
+      "loss": 4.0883,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9504659114509465e-05,
+      "loss": 4.0745,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949627316699894e-05,
+      "loss": 4.0687,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948788721948842e-05,
+      "loss": 4.0887,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94795012719779e-05,
+      "loss": 4.0631,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471131703271114e-05,
+      "loss": 4.071,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946274575576059e-05,
+      "loss": 4.084,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945435980825007e-05,
+      "loss": 4.0788,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944599023954328e-05,
+      "loss": 4.0664,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943760429203276e-05,
+      "loss": 4.0715,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429218344522236e-05,
+      "loss": 4.0631,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942083239701172e-05,
+      "loss": 4.0788,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94124464495012e-05,
+      "loss": 4.0853,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940406050199068e-05,
+      "loss": 4.078,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939567455448016e-05,
+      "loss": 4.079,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938728860696964e-05,
+      "loss": 4.0904,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937891903826285e-05,
+      "loss": 4.0916,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937054946955606e-05,
+      "loss": 4.0633,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936216352204554e-05,
+      "loss": 4.084,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935377757453502e-05,
+      "loss": 4.0781,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93453916270245e-05,
+      "loss": 4.0601,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933700567951398e-05,
+      "loss": 4.0813,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932861973200346e-05,
+      "loss": 4.075,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932023378449294e-05,
+      "loss": 4.087,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931186421578616e-05,
+      "loss": 4.0811,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303478268275637e-05,
+      "loss": 4.0636,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295092320765117e-05,
+      "loss": 4.0673,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9286706373254597e-05,
+      "loss": 4.0797,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278320425744076e-05,
+      "loss": 4.0758,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9269934478233556e-05,
+      "loss": 4.0788,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9261564909526766e-05,
+      "loss": 4.0799,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9253178962016245e-05,
+      "loss": 4.0716,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9244793014505725e-05,
+      "loss": 4.0727,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236407066995205e-05,
+      "loss": 4.0695,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9228021119484685e-05,
+      "loss": 4.0672,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9219635171974165e-05,
+      "loss": 4.0819,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211249224463645e-05,
+      "loss": 4.0729,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9202863276953125e-05,
+      "loss": 4.0816,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919449370824634e-05,
+      "loss": 4.0611,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918610776073582e-05,
+      "loss": 4.072,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91777218132253e-05,
+      "loss": 4.0803,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916935224451851e-05,
+      "loss": 4.0775,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916096629700799e-05,
+      "loss": 4.0739,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915258034949747e-05,
+      "loss": 4.0686,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914419440198695e-05,
+      "loss": 4.0749,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913580845447642e-05,
+      "loss": 4.0713,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91274225069659e-05,
+      "loss": 4.0604,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911903655945538e-05,
+      "loss": 4.0676,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911065061194486e-05,
+      "loss": 4.059,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910226466443434e-05,
+      "loss": 4.0802,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909389509572756e-05,
+      "loss": 4.0633,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908550914821704e-05,
+      "loss": 4.0734,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907712320070652e-05,
+      "loss": 4.0652,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068737253196e-05,
+      "loss": 4.0811,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906036768448921e-05,
+      "loss": 4.0678,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9051998115782424e-05,
+      "loss": 4.059,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90436121682719e-05,
+      "loss": 4.0782,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903522622076138e-05,
+      "loss": 4.0802,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902684027325086e-05,
+      "loss": 4.0732,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018470704544066e-05,
+      "loss": 4.075,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901008475703355e-05,
+      "loss": 4.0607,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900169880952303e-05,
+      "loss": 4.073,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899331286201251e-05,
+      "loss": 4.0738,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898492691450199e-05,
+      "loss": 4.0664,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897654096699147e-05,
+      "loss": 4.0628,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896815501948095e-05,
+      "loss": 4.0698,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895978545077416e-05,
+      "loss": 4.0747,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895139950326364e-05,
+      "loss": 4.0703,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894301355575312e-05,
+      "loss": 4.0499,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89346276082426e-05,
+      "loss": 4.0733,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892624166073208e-05,
+      "loss": 4.0599,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891785571322156e-05,
+      "loss": 4.0776,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890946976571104e-05,
+      "loss": 4.0743,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890108381820052e-05,
+      "loss": 4.0746,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889269787069e-05,
+      "loss": 4.055,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888432830198322e-05,
+      "loss": 4.0815,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88759423544727e-05,
+      "loss": 4.0718,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886755640696218e-05,
+      "loss": 4.0744,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885917045945166e-05,
+      "loss": 4.0688,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885078451194114e-05,
+      "loss": 4.0652,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8842414943234346e-05,
+      "loss": 4.0734,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8834028995723826e-05,
+      "loss": 4.0853,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8825643048213306e-05,
+      "loss": 4.0612,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817257100702786e-05,
+      "loss": 4.0715,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8808887531995995e-05,
+      "loss": 4.0774,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8800517963289204e-05,
+      "loss": 4.0617,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879214839458242e-05,
+      "loss": 4.0663,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87837624470719e-05,
+      "loss": 4.0686,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877537649956138e-05,
+      "loss": 4.0773,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876699055205086e-05,
+      "loss": 4.07,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875860460454034e-05,
+      "loss": 4.0668,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.08707857131958,
+      "eval_runtime": 294.7522,
+      "eval_samples_per_second": 1294.616,
+      "eval_steps_per_second": 40.458,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.875021865702982e-05,
+      "loss": 4.0529,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87418327095193e-05,
+      "loss": 4.0513,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873344676200878e-05,
+      "loss": 4.0664,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872506081449826e-05,
+      "loss": 4.0701,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871667486698773e-05,
+      "loss": 4.075,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870828891947721e-05,
+      "loss": 4.0627,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869990297196669e-05,
+      "loss": 4.0664,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869151702445617e-05,
+      "loss": 4.0598,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868313107694565e-05,
+      "loss": 4.0584,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867474512943514e-05,
+      "loss": 4.0668,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866635918192462e-05,
+      "loss": 4.0723,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86579732344141e-05,
+      "loss": 4.0697,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864960366570731e-05,
+      "loss": 4.0694,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864121771819679e-05,
+      "loss": 4.0564,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.863283177068627e-05,
+      "loss": 4.0611,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.862444582317575e-05,
+      "loss": 4.0459,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861605987566523e-05,
+      "loss": 4.0608,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.860767392815471e-05,
+      "loss": 4.0617,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859928798064419e-05,
+      "loss": 4.0601,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859090203313367e-05,
+      "loss": 4.0662,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858251608562314e-05,
+      "loss": 4.0747,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857413013811262e-05,
+      "loss": 4.0682,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856576056940584e-05,
+      "loss": 4.0656,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855737462189532e-05,
+      "loss": 4.0666,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85489886743848e-05,
+      "loss": 4.0588,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854060272687428e-05,
+      "loss": 4.0647,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853221677936376e-05,
+      "loss": 4.0634,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852383083185324e-05,
+      "loss": 4.0628,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851544488434272e-05,
+      "loss": 4.0558,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8507075315635926e-05,
+      "loss": 4.0531,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8498689368125406e-05,
+      "loss": 4.0573,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8490303420614886e-05,
+      "loss": 4.0699,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8481917473104366e-05,
+      "loss": 4.0667,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8473531525593846e-05,
+      "loss": 4.0642,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8465161956887055e-05,
+      "loss": 4.0565,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845677600937654e-05,
+      "loss": 4.0646,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844839006186602e-05,
+      "loss": 4.069,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84400041143555e-05,
+      "loss": 4.0506,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843161816684498e-05,
+      "loss": 4.0529,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842323221933446e-05,
+      "loss": 4.0472,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841484627182394e-05,
+      "loss": 4.0544,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840646032431342e-05,
+      "loss": 4.0665,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839809075560663e-05,
+      "loss": 4.0612,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838970480809611e-05,
+      "loss": 4.0557,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838131886058559e-05,
+      "loss": 4.067,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837293291307507e-05,
+      "loss": 4.0576,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836456334436828e-05,
+      "loss": 4.0649,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835617739685776e-05,
+      "loss": 4.064,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834779144934724e-05,
+      "loss": 4.031,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8339421880640455e-05,
+      "loss": 4.0764,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8331035933129935e-05,
+      "loss": 4.0541,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8322649985619415e-05,
+      "loss": 4.0675,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314264038108895e-05,
+      "loss": 4.0472,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8305878090598375e-05,
+      "loss": 4.054,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8297492143087855e-05,
+      "loss": 4.0484,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8289122574381064e-05,
+      "loss": 4.0518,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8280736626870544e-05,
+      "loss": 4.0521,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8272350679360024e-05,
+      "loss": 4.0474,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8263964731849504e-05,
+      "loss": 4.0674,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825557878433898e-05,
+      "loss": 4.0481,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8247192836828464e-05,
+      "loss": 4.0413,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8238806889317944e-05,
+      "loss": 4.0627,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8230420941807424e-05,
+      "loss": 4.0389,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8222034994296903e-05,
+      "loss": 4.0449,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821366542559011e-05,
+      "loss": 4.0583,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820527947807959e-05,
+      "loss": 4.0528,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819690990937281e-05,
+      "loss": 4.044,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818852396186228e-05,
+      "loss": 4.0433,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818013801435176e-05,
+      "loss": 4.04,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817175206684124e-05,
+      "loss": 4.0545,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816336611933072e-05,
+      "loss": 4.061,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81549801718202e-05,
+      "loss": 4.052,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814659422430968e-05,
+      "loss": 4.0546,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813820827679916e-05,
+      "loss": 4.0623,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812983870809238e-05,
+      "loss": 4.0689,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8121469139385586e-05,
+      "loss": 4.0364,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8113083191875066e-05,
+      "loss": 4.0606,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8104697244364546e-05,
+      "loss": 4.0499,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8096311296854026e-05,
+      "loss": 4.0383,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8087925349343506e-05,
+      "loss": 4.0514,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8079539401832986e-05,
+      "loss": 4.0596,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8071153454322466e-05,
+      "loss": 4.0578,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8062783885615675e-05,
+      "loss": 4.0629,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8054397938105155e-05,
+      "loss": 4.0388,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8046011990594635e-05,
+      "loss": 4.0407,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8037626043084115e-05,
+      "loss": 4.0567,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80292400955736e-05,
+      "loss": 4.0486,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802087052686681e-05,
+      "loss": 4.0514,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801248457935629e-05,
+      "loss": 4.0585,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800409863184577e-05,
+      "loss": 4.0465,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799571268433525e-05,
+      "loss": 4.0481,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798732673682473e-05,
+      "loss": 4.0476,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797895716811794e-05,
+      "loss": 4.0421,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797057122060742e-05,
+      "loss": 4.0568,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79621852730969e-05,
+      "loss": 4.0516,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795379932558638e-05,
+      "loss": 4.0531,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794541337807586e-05,
+      "loss": 4.0425,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793704380936907e-05,
+      "loss": 4.0421,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792865786185855e-05,
+      "loss": 4.0571,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920271914348035e-05,
+      "loss": 4.053,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911885966837515e-05,
+      "loss": 4.0517,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7903500019326995e-05,
+      "loss": 4.0481,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789511407181647e-05,
+      "loss": 4.0485,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788672812430595e-05,
+      "loss": 4.0422,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787834217679543e-05,
+      "loss": 4.0429,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786998898689237e-05,
+      "loss": 4.0392,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786160303938185e-05,
+      "loss": 4.0369,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785321709187133e-05,
+      "loss": 4.0575,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784483114436081e-05,
+      "loss": 4.0433,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7836445196850286e-05,
+      "loss": 4.0496,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782805924933977e-05,
+      "loss": 4.0398,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781967330182925e-05,
+      "loss": 4.0536,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781128735431873e-05,
+      "loss": 4.0454,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780291778561194e-05,
+      "loss": 4.0364,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779453183810142e-05,
+      "loss": 4.0563,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778616226939464e-05,
+      "loss": 4.0535,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777777632188411e-05,
+      "loss": 4.0511,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776939037437359e-05,
+      "loss": 4.0504,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776100442686307e-05,
+      "loss": 4.0419,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775261847935255e-05,
+      "loss": 4.0477,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774424891064576e-05,
+      "loss": 4.0533,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773586296313524e-05,
+      "loss": 4.041,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772747701562473e-05,
+      "loss": 4.0406,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771909106811421e-05,
+      "loss": 4.0469,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771070512060369e-05,
+      "loss": 4.0502,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7702335551896896e-05,
+      "loss": 4.0456,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7693949604386376e-05,
+      "loss": 4.0313,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7685563656875856e-05,
+      "loss": 4.0513,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7677177709365336e-05,
+      "loss": 4.0337,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7668791761854816e-05,
+      "loss": 4.0562,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7660405814344296e-05,
+      "loss": 4.0462,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7652019866833776e-05,
+      "loss": 4.0561,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7643633919323256e-05,
+      "loss": 4.0269,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7635247971812735e-05,
+      "loss": 4.0608,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7626862024302215e-05,
+      "loss": 4.0471,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7618492455595425e-05,
+      "loss": 4.0522,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761010650808491e-05,
+      "loss": 4.0468,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760172056057439e-05,
+      "loss": 4.0381,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75933509918676e-05,
+      "loss": 4.0552,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758496504435708e-05,
+      "loss": 4.0608,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757657909684656e-05,
+      "loss": 4.0414,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756819314933604e-05,
+      "loss": 4.0445,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755980720182552e-05,
+      "loss": 4.0548,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7551421254315e-05,
+      "loss": 4.0403,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754303530680447e-05,
+      "loss": 4.0435,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753464935929395e-05,
+      "loss": 4.0486,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752627979058717e-05,
+      "loss": 4.05,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751789384307665e-05,
+      "loss": 4.0512,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.750950789556613e-05,
+      "loss": 4.041,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.071474075317383,
+      "eval_runtime": 292.6612,
+      "eval_samples_per_second": 1303.866,
+      "eval_steps_per_second": 40.747,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750112194805561e-05,
+      "loss": 4.0321,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7492752379348825e-05,
+      "loss": 4.0294,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74843664318383e-05,
+      "loss": 4.0427,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7475996863131514e-05,
+      "loss": 4.0488,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7467610915620994e-05,
+      "loss": 4.0519,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7459224968110474e-05,
+      "loss": 4.0416,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745083902059995e-05,
+      "loss": 4.0434,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744245307308943e-05,
+      "loss": 4.0358,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743406712557891e-05,
+      "loss": 4.0391,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742568117806839e-05,
+      "loss": 4.0424,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74173116093616e-05,
+      "loss": 4.0484,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740892566185108e-05,
+      "loss": 4.0473,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74005560931443e-05,
+      "loss": 4.0522,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739217014563377e-05,
+      "loss": 4.0328,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738378419812325e-05,
+      "loss": 4.0426,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737539825061273e-05,
+      "loss": 4.02,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736701230310221e-05,
+      "loss": 4.0372,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735862635559169e-05,
+      "loss": 4.0391,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735024040808117e-05,
+      "loss": 4.0373,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734185446057065e-05,
+      "loss": 4.041,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733346851306013e-05,
+      "loss": 4.0513,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732509894435334e-05,
+      "loss": 4.0455,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731671299684282e-05,
+      "loss": 4.047,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.73083270493323e-05,
+      "loss": 4.0411,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729994110182178e-05,
+      "loss": 4.0409,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729155515431127e-05,
+      "loss": 4.0418,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728316920680075e-05,
+      "loss": 4.0398,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727478325929023e-05,
+      "loss": 4.0416,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726639731177971e-05,
+      "loss": 4.0339,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725801136426919e-05,
+      "loss": 4.0312,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7249658174366125e-05,
+      "loss": 4.0324,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7241272226855605e-05,
+      "loss": 4.0512,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7232886279345085e-05,
+      "loss": 4.0485,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7224500331834565e-05,
+      "loss": 4.0379,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7216114384324045e-05,
+      "loss": 4.0386,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7207744815617254e-05,
+      "loss": 4.0359,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7199358868106734e-05,
+      "loss": 4.0495,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719097292059622e-05,
+      "loss": 4.032,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71825869730857e-05,
+      "loss": 4.0298,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717420102557518e-05,
+      "loss": 4.026,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716581507806466e-05,
+      "loss": 4.0338,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7157429130554134e-05,
+      "loss": 4.0412,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7149043183043614e-05,
+      "loss": 4.0391,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714067361433683e-05,
+      "loss": 4.0368,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713228766682631e-05,
+      "loss": 4.0443,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712390171931578e-05,
+      "loss": 4.0334,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711551577180526e-05,
+      "loss": 4.0444,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710714620309848e-05,
+      "loss": 4.0403,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709876025558796e-05,
+      "loss": 4.0121,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709037430807744e-05,
+      "loss": 4.0522,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708198836056692e-05,
+      "loss": 4.0316,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70736024130564e-05,
+      "loss": 4.0468,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706523284434961e-05,
+      "loss": 4.0287,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705684689683909e-05,
+      "loss": 4.0303,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704846094932857e-05,
+      "loss": 4.0263,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704009138062178e-05,
+      "loss": 4.0334,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7031705433111256e-05,
+      "loss": 4.0306,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7023319485600736e-05,
+      "loss": 4.0236,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7014933538090216e-05,
+      "loss": 4.045,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7006547590579696e-05,
+      "loss": 4.0318,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6998161643069176e-05,
+      "loss": 4.0153,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698979207436239e-05,
+      "loss": 4.0428,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698140612685187e-05,
+      "loss": 4.0218,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.697302017934135e-05,
+      "loss": 4.0213,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696463423183083e-05,
+      "loss": 4.0367,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695624828432031e-05,
+      "loss": 4.0319,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694786233680979e-05,
+      "loss": 4.0215,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693947638929927e-05,
+      "loss": 4.0223,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693109044178875e-05,
+      "loss": 4.0206,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692272087308196e-05,
+      "loss": 4.0324,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691433492557144e-05,
+      "loss": 4.0382,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690594897806092e-05,
+      "loss": 4.0336,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68975630305504e-05,
+      "loss": 4.0356,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688917708303988e-05,
+      "loss": 4.0382,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68808075143331e-05,
+      "loss": 4.0459,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6872437945626306e-05,
+      "loss": 4.0244,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6864051998115786e-05,
+      "loss": 4.0335,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6855666050605266e-05,
+      "loss": 4.0272,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6847280103094746e-05,
+      "loss": 4.0197,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6838894155584226e-05,
+      "loss": 4.0296,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6830540965681164e-05,
+      "loss": 4.0404,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6822155018170644e-05,
+      "loss": 4.0351,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6813769070660124e-05,
+      "loss": 4.0434,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6805383123149604e-05,
+      "loss": 4.0122,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6796997175639084e-05,
+      "loss": 4.0259,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6788611228128564e-05,
+      "loss": 4.0333,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678022528061805e-05,
+      "loss": 4.0323,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677183933310753e-05,
+      "loss": 4.0235,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676345338559701e-05,
+      "loss": 4.0403,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675506743808649e-05,
+      "loss": 4.029,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674668149057597e-05,
+      "loss": 4.03,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673831192186918e-05,
+      "loss": 4.019,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672992597435866e-05,
+      "loss": 4.0229,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672154002684814e-05,
+      "loss": 4.0358,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671317045814135e-05,
+      "loss": 4.0308,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670478451063083e-05,
+      "loss": 4.0333,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669639856312031e-05,
+      "loss": 4.0243,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668801261560979e-05,
+      "loss": 4.0196,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667962666809927e-05,
+      "loss": 4.0342,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667124072058875e-05,
+      "loss": 4.035,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666285477307823e-05,
+      "loss": 4.032,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665446882556771e-05,
+      "loss": 4.0268,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664608287805719e-05,
+      "loss": 4.0277,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663769693054667e-05,
+      "loss": 4.0238,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662931098303615e-05,
+      "loss": 4.0256,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662092503552563e-05,
+      "loss": 4.0164,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661255546681884e-05,
+      "loss": 4.0184,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660416951930832e-05,
+      "loss": 4.0285,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65957835717978e-05,
+      "loss": 4.0298,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658739762428728e-05,
+      "loss": 4.0218,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657901167677676e-05,
+      "loss": 4.0248,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6570642108069966e-05,
+      "loss": 4.0297,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.656227253936318e-05,
+      "loss": 4.0267,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.655388659185266e-05,
+      "loss": 4.014,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.654550064434214e-05,
+      "loss": 4.0354,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653711469683162e-05,
+      "loss": 4.0337,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65287287493211e-05,
+      "loss": 4.0335,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652034280181058e-05,
+      "loss": 4.0258,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651195685430006e-05,
+      "loss": 4.0204,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650357090678954e-05,
+      "loss": 4.0267,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649520133808275e-05,
+      "loss": 4.0374,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648681539057223e-05,
+      "loss": 4.0202,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647842944306171e-05,
+      "loss": 4.0209,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647004349555119e-05,
+      "loss": 4.0202,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646165754804067e-05,
+      "loss": 4.0327,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645327160053015e-05,
+      "loss": 4.0262,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644488565301963e-05,
+      "loss": 4.0082,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643649970550911e-05,
+      "loss": 4.0306,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642811375799859e-05,
+      "loss": 4.0162,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641972781048807e-05,
+      "loss": 4.0359,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641135824178128e-05,
+      "loss": 4.0269,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640297229427076e-05,
+      "loss": 4.034,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639458634676024e-05,
+      "loss": 4.0106,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638620039924972e-05,
+      "loss": 4.0372,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63778144517392e-05,
+      "loss": 4.0291,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636944488303241e-05,
+      "loss": 4.0293,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636105893552189e-05,
+      "loss": 4.0293,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6352672988011375e-05,
+      "loss": 4.0165,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6344287040500855e-05,
+      "loss": 4.0371,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6335901092990335e-05,
+      "loss": 4.0352,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6327515145479815e-05,
+      "loss": 4.025,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6319129197969295e-05,
+      "loss": 4.0249,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6310743250458775e-05,
+      "loss": 4.0323,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6302373681751984e-05,
+      "loss": 4.0243,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629400411304519e-05,
+      "loss": 4.0182,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628561816553467e-05,
+      "loss": 4.0299,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627723221802415e-05,
+      "loss": 4.0299,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626884627051363e-05,
+      "loss": 4.0333,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626046032300311e-05,
+      "loss": 4.0223,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.058262348175049,
+      "eval_runtime": 291.8569,
+      "eval_samples_per_second": 1307.459,
+      "eval_steps_per_second": 40.859,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625207437549259e-05,
+      "loss": 4.0061,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624368842798207e-05,
+      "loss": 4.0101,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623530248047155e-05,
+      "loss": 4.0234,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622691653296104e-05,
+      "loss": 4.0274,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621853058545052e-05,
+      "loss": 4.0321,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621014463793999e-05,
+      "loss": 4.0221,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620175869042947e-05,
+      "loss": 4.0242,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619337274291895e-05,
+      "loss": 4.0201,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618498679540843e-05,
+      "loss": 4.0175,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.617660084789791e-05,
+      "loss": 4.0194,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.616821490038739e-05,
+      "loss": 4.0308,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615982895287687e-05,
+      "loss": 4.0305,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615144300536635e-05,
+      "loss": 4.0306,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.614305705785583e-05,
+      "loss": 4.0111,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613467111034531e-05,
+      "loss": 4.0225,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612628516283479e-05,
+      "loss": 3.9992,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611791559412801e-05,
+      "loss": 4.0225,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610952964661749e-05,
+      "loss": 4.0232,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610114369910697e-05,
+      "loss": 4.0153,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609275775159645e-05,
+      "loss": 4.0224,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608437180408593e-05,
+      "loss": 4.0298,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607598585657541e-05,
+      "loss": 4.0245,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606759990906488e-05,
+      "loss": 4.0318,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605921396155436e-05,
+      "loss": 4.0224,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6050844392847576e-05,
+      "loss": 4.0215,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6042474824140786e-05,
+      "loss": 4.0192,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6034088876630266e-05,
+      "loss": 4.0172,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6025702929119745e-05,
+      "loss": 4.0241,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6017316981609225e-05,
+      "loss": 4.0177,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6008931034098705e-05,
+      "loss": 4.0102,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600056146539192e-05,
+      "loss": 4.0134,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59921755178814e-05,
+      "loss": 4.0272,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598378957037088e-05,
+      "loss": 4.0286,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5975403622860354e-05,
+      "loss": 4.0222,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5967017675349834e-05,
+      "loss": 4.0183,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595864810664305e-05,
+      "loss": 4.0217,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595026215913253e-05,
+      "loss": 4.029,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5941876211622e-05,
+      "loss": 4.0086,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593349026411148e-05,
+      "loss": 4.0133,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592510431660096e-05,
+      "loss": 4.0114,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591671836909044e-05,
+      "loss": 4.0141,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590833242157993e-05,
+      "loss": 4.0234,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589994647406941e-05,
+      "loss": 4.0139,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589157690536262e-05,
+      "loss": 4.0231,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58831909578521e-05,
+      "loss": 4.023,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587480501034158e-05,
+      "loss": 4.0191,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.586641906283106e-05,
+      "loss": 4.0202,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585804949412427e-05,
+      "loss": 4.025,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584966354661375e-05,
+      "loss": 3.993,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584127759910323e-05,
+      "loss": 4.0299,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583289165159271e-05,
+      "loss": 4.0146,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582450570408219e-05,
+      "loss": 4.0275,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58161361353754e-05,
+      "loss": 4.01,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5807750187864884e-05,
+      "loss": 4.0137,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579938061915809e-05,
+      "loss": 4.0027,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579099467164757e-05,
+      "loss": 4.0114,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578260872413705e-05,
+      "loss": 4.0182,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.577422277662653e-05,
+      "loss": 3.9999,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576583682911601e-05,
+      "loss": 4.0269,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575745088160549e-05,
+      "loss": 4.0144,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574906493409497e-05,
+      "loss": 3.9963,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574067898658445e-05,
+      "loss": 4.0224,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573230941787766e-05,
+      "loss": 4.0053,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572392347036714e-05,
+      "loss": 4.0015,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571555390166035e-05,
+      "loss": 4.0172,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.570716795414983e-05,
+      "loss": 4.0153,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569878200663932e-05,
+      "loss": 4.0028,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56903960591288e-05,
+      "loss": 4.0045,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.568201011161828e-05,
+      "loss": 4.0013,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567362416410776e-05,
+      "loss": 4.0141,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566523821659724e-05,
+      "loss": 4.0188,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5656868647890446e-05,
+      "loss": 4.0109,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5648482700379926e-05,
+      "loss": 4.0184,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5640096752869406e-05,
+      "loss": 4.0179,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5631710805358886e-05,
+      "loss": 4.0256,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5623341236652095e-05,
+      "loss": 4.0065,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5614955289141575e-05,
+      "loss": 4.0188,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5606569341631055e-05,
+      "loss": 4.0056,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5598183394120535e-05,
+      "loss": 4.0018,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5589797446610015e-05,
+      "loss": 4.0134,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558142787790323e-05,
+      "loss": 4.0163,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557304193039271e-05,
+      "loss": 4.0205,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556465598288219e-05,
+      "loss": 4.0247,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55562864141754e-05,
+      "loss": 3.9903,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554790046666488e-05,
+      "loss": 4.0067,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553951451915436e-05,
+      "loss": 4.0134,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553112857164383e-05,
+      "loss": 4.0151,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552274262413331e-05,
+      "loss": 4.007,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.551435667662279e-05,
+      "loss": 4.0206,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550597072911227e-05,
+      "loss": 4.0133,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549758478160175e-05,
+      "loss": 4.0096,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548921521289497e-05,
+      "loss": 4.0048,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548082926538445e-05,
+      "loss": 4.0032,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547244331787393e-05,
+      "loss": 4.0157,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546405737036341e-05,
+      "loss": 4.0144,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545568780165662e-05,
+      "loss": 4.0118,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54473018541461e-05,
+      "loss": 4.0071,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543891590663558e-05,
+      "loss": 4.005,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5430546337928787e-05,
+      "loss": 4.0169,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5422160390418267e-05,
+      "loss": 4.0159,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5413774442907746e-05,
+      "loss": 4.0114,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5405388495397226e-05,
+      "loss": 4.008,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5397002547886706e-05,
+      "loss": 4.01,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538861660037619e-05,
+      "loss": 4.0091,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538023065286567e-05,
+      "loss": 4.0021,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537184470535515e-05,
+      "loss": 4.0005,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.536347513664836e-05,
+      "loss": 4.0019,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.535508918913784e-05,
+      "loss": 4.0145,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534670324162732e-05,
+      "loss": 4.0067,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53383172941168e-05,
+      "loss": 4.0081,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532994772541001e-05,
+      "loss": 4.0011,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532156177789949e-05,
+      "loss": 4.0101,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531317583038897e-05,
+      "loss": 4.0124,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530478988287845e-05,
+      "loss": 4.0008,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529642031417166e-05,
+      "loss": 4.0113,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528803436666115e-05,
+      "loss": 4.0127,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527964841915063e-05,
+      "loss": 4.0161,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527126247164011e-05,
+      "loss": 4.009,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5262892902933316e-05,
+      "loss": 4.0073,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5254523334226525e-05,
+      "loss": 4.008,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5246137386716005e-05,
+      "loss": 4.0221,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5237751439205485e-05,
+      "loss": 4.0027,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5229365491694965e-05,
+      "loss": 4.0016,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5220979544184445e-05,
+      "loss": 4.0,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5212609975477654e-05,
+      "loss": 4.0134,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5204224027967134e-05,
+      "loss": 4.0111,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5195838080456614e-05,
+      "loss": 3.9945,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51874521329461e-05,
+      "loss": 4.0071,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517906618543558e-05,
+      "loss": 3.9959,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517068023792506e-05,
+      "loss": 4.0182,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516229429041454e-05,
+      "loss": 4.0136,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515390834290402e-05,
+      "loss": 4.013,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5145522395393493e-05,
+      "loss": 3.9979,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513715282668671e-05,
+      "loss": 4.0127,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512876687917619e-05,
+      "loss": 4.0151,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512038093166567e-05,
+      "loss": 4.0089,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511199498415514e-05,
+      "loss": 4.0119,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510362541544836e-05,
+      "loss": 3.9993,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509523946793784e-05,
+      "loss": 4.0217,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508685352042732e-05,
+      "loss": 4.0169,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5078483951720534e-05,
+      "loss": 4.009,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5070098004210014e-05,
+      "loss": 4.0045,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5061712056699494e-05,
+      "loss": 4.016,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505332610918897e-05,
+      "loss": 4.0043,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504494016167845e-05,
+      "loss": 4.0015,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503655421416793e-05,
+      "loss": 4.013,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502818464546114e-05,
+      "loss": 4.0108,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5019798697950616e-05,
+      "loss": 4.0154,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5011412750440096e-05,
+      "loss": 4.0048,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0469231605529785,
+      "eval_runtime": 289.8326,
+      "eval_samples_per_second": 1316.591,
+      "eval_steps_per_second": 41.144,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5003026802929576e-05,
+      "loss": 3.9987,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4994640855419056e-05,
+      "loss": 3.9891,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4986254907908536e-05,
+      "loss": 4.0047,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4977868960398016e-05,
+      "loss": 4.0104,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.49694830128875e-05,
+      "loss": 4.0106,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.496111344418071e-05,
+      "loss": 4.0054,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.495272749667019e-05,
+      "loss": 4.0129,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.494434154915967e-05,
+      "loss": 3.9999,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.493595560164915e-05,
+      "loss": 3.9975,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492756965413863e-05,
+      "loss": 4.0006,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491920008543184e-05,
+      "loss": 4.0156,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491081413792132e-05,
+      "loss": 4.0119,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.490244456921453e-05,
+      "loss": 4.0126,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.489405862170401e-05,
+      "loss": 3.997,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488567267419349e-05,
+      "loss": 4.0068,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487728672668297e-05,
+      "loss": 3.9796,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4868900779172456e-05,
+      "loss": 4.0047,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4860531210465665e-05,
+      "loss": 4.0041,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4852145262955145e-05,
+      "loss": 3.996,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4843759315444625e-05,
+      "loss": 4.007,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4835373367934105e-05,
+      "loss": 4.01,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4826987420423585e-05,
+      "loss": 4.0076,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4818617851716794e-05,
+      "loss": 4.0145,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4810248283010003e-05,
+      "loss": 4.0058,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.480186233549948e-05,
+      "loss": 4.0042,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.479347638798896e-05,
+      "loss": 4.0025,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.478509044047844e-05,
+      "loss": 3.9988,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.477670449296792e-05,
+      "loss": 4.0062,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476831854545741e-05,
+      "loss": 4.0009,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.475993259794689e-05,
+      "loss": 3.9906,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.475154665043637e-05,
+      "loss": 3.9982,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.474316070292585e-05,
+      "loss": 4.007,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.473479113421906e-05,
+      "loss": 4.0086,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.472640518670854e-05,
+      "loss": 4.0077,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471801923919802e-05,
+      "loss": 4.0024,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470964967049123e-05,
+      "loss": 4.0047,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470126372298071e-05,
+      "loss": 4.0107,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.469287777547019e-05,
+      "loss": 3.9919,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468449182795967e-05,
+      "loss": 3.9954,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467610588044915e-05,
+      "loss": 3.9943,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466771993293863e-05,
+      "loss": 3.9946,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465933398542811e-05,
+      "loss": 4.0033,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465094803791759e-05,
+      "loss": 4.0014,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464256209040707e-05,
+      "loss": 4.0056,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463417614289655e-05,
+      "loss": 4.0044,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462579019538603e-05,
+      "loss": 4.0009,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461740424787551e-05,
+      "loss": 4.0024,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4609034679168717e-05,
+      "loss": 4.0134,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4600648731658197e-05,
+      "loss": 3.9764,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4592262784147676e-05,
+      "loss": 4.0133,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4583876836637156e-05,
+      "loss": 3.9955,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4575507267930366e-05,
+      "loss": 4.0073,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4567121320419845e-05,
+      "loss": 3.9933,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455873537290933e-05,
+      "loss": 3.9952,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455034942539881e-05,
+      "loss": 3.9903,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.454197985669202e-05,
+      "loss": 3.993,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45335939091815e-05,
+      "loss": 4.0027,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.452520796167098e-05,
+      "loss": 3.985,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.451683839296419e-05,
+      "loss": 4.0113,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450845244545367e-05,
+      "loss": 3.9973,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450006649794315e-05,
+      "loss": 3.9777,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449168055043263e-05,
+      "loss": 4.0061,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448329460292211e-05,
+      "loss": 3.9867,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.447490865541159e-05,
+      "loss": 3.9873,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.446652270790107e-05,
+      "loss": 3.9984,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445813676039055e-05,
+      "loss": 3.997,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4449767191683766e-05,
+      "loss": 3.9858,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4441381244173246e-05,
+      "loss": 3.993,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4432995296662726e-05,
+      "loss": 3.984,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4424609349152206e-05,
+      "loss": 3.9927,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4416239780445415e-05,
+      "loss": 4.0033,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4407853832934895e-05,
+      "loss": 3.9982,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4399467885424375e-05,
+      "loss": 4.0022,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4391081937913855e-05,
+      "loss": 4.0011,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4382695990403335e-05,
+      "loss": 4.0103,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4374326421696544e-05,
+      "loss": 3.9927,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4365940474186024e-05,
+      "loss": 4.0041,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4357554526675504e-05,
+      "loss": 3.9863,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4349168579164984e-05,
+      "loss": 3.9847,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4340782631654464e-05,
+      "loss": 3.9964,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4332396684143943e-05,
+      "loss": 3.9964,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4324010736633423e-05,
+      "loss": 4.0059,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4315624789122903e-05,
+      "loss": 4.0051,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.430723884161238e-05,
+      "loss": 3.9837,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429886927290559e-05,
+      "loss": 3.9879,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429048332539507e-05,
+      "loss": 3.9939,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.428209737788455e-05,
+      "loss": 4.0012,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.427371143037403e-05,
+      "loss": 3.9887,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.426535824047098e-05,
+      "loss": 4.0034,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.425697229296046e-05,
+      "loss": 4.0027,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424858634544994e-05,
+      "loss": 3.9888,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424020039793942e-05,
+      "loss": 3.9898,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.42318144504289e-05,
+      "loss": 3.9894,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422342850291838e-05,
+      "loss": 3.9968,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.421504255540786e-05,
+      "loss": 4.001,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.420665660789734e-05,
+      "loss": 3.9969,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.419827066038682e-05,
+      "loss": 3.9921,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418991747048376e-05,
+      "loss": 3.9846,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4181531522973235e-05,
+      "loss": 4.004,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4173145575462715e-05,
+      "loss": 3.9993,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4164759627952195e-05,
+      "loss": 4.0009,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4156373680441675e-05,
+      "loss": 3.9889,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4147987732931155e-05,
+      "loss": 3.993,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413960178542064e-05,
+      "loss": 3.9904,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413121583791012e-05,
+      "loss": 3.9879,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.412284626920333e-05,
+      "loss": 3.987,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.411447670049654e-05,
+      "loss": 3.9842,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.410609075298602e-05,
+      "loss": 3.9934,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40977048054755e-05,
+      "loss": 3.9978,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408931885796498e-05,
+      "loss": 3.9852,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408093291045446e-05,
+      "loss": 3.9887,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.407254696294394e-05,
+      "loss": 3.9925,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.406416101543342e-05,
+      "loss": 3.9967,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40557750679229e-05,
+      "loss": 3.9837,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.404740549921611e-05,
+      "loss": 4.0003,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4039019551705595e-05,
+      "loss": 3.9922,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4030633604195075e-05,
+      "loss": 4.0042,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4022264035488284e-05,
+      "loss": 3.9938,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4013878087977764e-05,
+      "loss": 3.9905,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4005492140467244e-05,
+      "loss": 3.9942,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3997106192956724e-05,
+      "loss": 4.0019,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3988720245446204e-05,
+      "loss": 3.9845,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398035067673941e-05,
+      "loss": 3.9893,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.397196472922889e-05,
+      "loss": 3.9879,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396357878171837e-05,
+      "loss": 3.9931,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.395519283420785e-05,
+      "loss": 3.99,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394680688669733e-05,
+      "loss": 3.9824,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393843731799055e-05,
+      "loss": 3.9883,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393005137048003e-05,
+      "loss": 3.9839,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392166542296951e-05,
+      "loss": 4.0048,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391327947545899e-05,
+      "loss": 3.9932,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390489352794847e-05,
+      "loss": 3.9993,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389650758043795e-05,
+      "loss": 3.9811,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388812163292742e-05,
+      "loss": 3.9985,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.38797356854169e-05,
+      "loss": 4.0008,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387136611671012e-05,
+      "loss": 3.995,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.386298016919959e-05,
+      "loss": 3.9907,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.385459422168907e-05,
+      "loss": 3.9831,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.384622465298229e-05,
+      "loss": 4.0089,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.383783870547177e-05,
+      "loss": 4.0017,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382945275796125e-05,
+      "loss": 3.9939,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382106681045073e-05,
+      "loss": 3.9876,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.381268086294021e-05,
+      "loss": 4.0032,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.380429491542969e-05,
+      "loss": 3.9916,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3795908967919167e-05,
+      "loss": 3.9832,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3787523020408647e-05,
+      "loss": 4.0032,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3779137072898126e-05,
+      "loss": 3.9903,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3770751125387606e-05,
+      "loss": 4.0022,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3762365177877086e-05,
+      "loss": 3.9912,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.038114070892334,
+      "eval_runtime": 297.5213,
+      "eval_samples_per_second": 1282.567,
+      "eval_steps_per_second": 40.081,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3753979230366566e-05,
+      "loss": 4.0002,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3745593282856046e-05,
+      "loss": 3.9714,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3737207335345526e-05,
+      "loss": 3.9875,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3728821387835006e-05,
+      "loss": 4.0007,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372043544032448e-05,
+      "loss": 3.9917,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3712049492813966e-05,
+      "loss": 3.9917,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3703663545303446e-05,
+      "loss": 3.9951,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3695277597792926e-05,
+      "loss": 3.9857,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3686891650282406e-05,
+      "loss": 3.9822,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3678505702771886e-05,
+      "loss": 3.9889,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3670119755261366e-05,
+      "loss": 3.9984,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3661733807750846e-05,
+      "loss": 3.9932,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3653347860240326e-05,
+      "loss": 4.0048,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3644961912729806e-05,
+      "loss": 3.9792,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3636575965219286e-05,
+      "loss": 3.9858,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3628190017708766e-05,
+      "loss": 3.9695,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3619820449001975e-05,
+      "loss": 3.986,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3611434501491455e-05,
+      "loss": 3.9917,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3603048553980935e-05,
+      "loss": 3.9833,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3594662606470415e-05,
+      "loss": 3.9824,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3586276658959895e-05,
+      "loss": 4.0011,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3577890711449375e-05,
+      "loss": 3.9911,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3569504763938855e-05,
+      "loss": 3.9977,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3561118816428334e-05,
+      "loss": 3.9924,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3552732868917814e-05,
+      "loss": 3.9924,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354437967901476e-05,
+      "loss": 3.9847,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353599373150424e-05,
+      "loss": 3.9833,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352760778399372e-05,
+      "loss": 3.9923,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35192218364832e-05,
+      "loss": 3.9868,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351083588897267e-05,
+      "loss": 3.9725,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350244994146215e-05,
+      "loss": 3.9862,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349406399395163e-05,
+      "loss": 3.9899,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348567804644111e-05,
+      "loss": 3.9944,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.34772920989306e-05,
+      "loss": 3.9888,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3468938909027544e-05,
+      "loss": 3.9906,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346056934032075e-05,
+      "loss": 3.9942,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345218339281023e-05,
+      "loss": 3.9961,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344379744529971e-05,
+      "loss": 3.9704,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343541149778919e-05,
+      "loss": 3.9816,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342702555027867e-05,
+      "loss": 3.9785,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3418639602768146e-05,
+      "loss": 3.9814,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3410253655257626e-05,
+      "loss": 3.9926,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3401867707747106e-05,
+      "loss": 3.9907,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3393481760236586e-05,
+      "loss": 3.9877,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3385095812726066e-05,
+      "loss": 3.9891,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337670986521555e-05,
+      "loss": 3.9833,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336832391770503e-05,
+      "loss": 3.9878,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335993797019451e-05,
+      "loss": 3.996,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335156840148772e-05,
+      "loss": 3.9626,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33431824539772e-05,
+      "loss": 3.9963,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333479650646668e-05,
+      "loss": 3.9854,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332641055895616e-05,
+      "loss": 3.9921,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331804099024937e-05,
+      "loss": 3.9789,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330965504273885e-05,
+      "loss": 3.9796,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330126909522833e-05,
+      "loss": 3.9769,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329289952652154e-05,
+      "loss": 3.9758,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.328451357901102e-05,
+      "loss": 3.9878,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3276127631500506e-05,
+      "loss": 3.9684,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3267741683989986e-05,
+      "loss": 3.9962,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3259355736479466e-05,
+      "loss": 3.9826,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3250969788968946e-05,
+      "loss": 3.9636,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3242583841458426e-05,
+      "loss": 3.9915,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3234197893947906e-05,
+      "loss": 3.9694,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3225828325241115e-05,
+      "loss": 3.976,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3217442377730595e-05,
+      "loss": 3.9817,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3209056430220075e-05,
+      "loss": 3.9875,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3200670482709555e-05,
+      "loss": 3.9684,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.319228453519903e-05,
+      "loss": 3.9829,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3183914966492244e-05,
+      "loss": 3.965,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3175529018981724e-05,
+      "loss": 3.9778,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3167143071471204e-05,
+      "loss": 3.991,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315877350276442e-05,
+      "loss": 3.9795,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31503875552539e-05,
+      "loss": 3.9875,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314200160774338e-05,
+      "loss": 3.9826,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.313361566023286e-05,
+      "loss": 3.9986,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.312522971272233e-05,
+      "loss": 3.9782,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.311686014401555e-05,
+      "loss": 3.9873,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310847419650503e-05,
+      "loss": 3.9743,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31000882489945e-05,
+      "loss": 3.9699,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309170230148398e-05,
+      "loss": 3.9836,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30833327327772e-05,
+      "loss": 3.9785,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3074963164070414e-05,
+      "loss": 3.9883,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3066577216559894e-05,
+      "loss": 3.9956,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30582076478531e-05,
+      "loss": 3.9743,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304982170034258e-05,
+      "loss": 3.971,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304143575283206e-05,
+      "loss": 3.9775,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303304980532154e-05,
+      "loss": 3.9856,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302466385781102e-05,
+      "loss": 3.9754,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301629428910423e-05,
+      "loss": 3.9848,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300790834159371e-05,
+      "loss": 3.9863,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299952239408319e-05,
+      "loss": 3.9753,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299113644657267e-05,
+      "loss": 3.9781,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298275049906215e-05,
+      "loss": 3.969,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297436455155163e-05,
+      "loss": 3.9898,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296597860404111e-05,
+      "loss": 3.9823,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295759265653059e-05,
+      "loss": 3.9839,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294920670902007e-05,
+      "loss": 3.9757,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294082076150955e-05,
+      "loss": 3.972,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293243481399903e-05,
+      "loss": 3.989,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292404886648851e-05,
+      "loss": 3.9846,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291567929778172e-05,
+      "loss": 3.9828,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29072933502712e-05,
+      "loss": 3.9786,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289890740276068e-05,
+      "loss": 3.9749,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289052145525016e-05,
+      "loss": 3.9791,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.288213550773964e-05,
+      "loss": 3.9757,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287374956022912e-05,
+      "loss": 3.9687,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2865379991522336e-05,
+      "loss": 3.9721,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2856994044011816e-05,
+      "loss": 3.9778,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2848608096501296e-05,
+      "loss": 3.9848,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2840222148990776e-05,
+      "loss": 3.9682,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2831836201480256e-05,
+      "loss": 3.9777,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2823450253969736e-05,
+      "loss": 3.9729,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2815064306459216e-05,
+      "loss": 3.9859,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2806694737752425e-05,
+      "loss": 3.9719,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2798308790241905e-05,
+      "loss": 3.9794,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2789922842731385e-05,
+      "loss": 3.9813,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2781536895220865e-05,
+      "loss": 3.9911,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2773167326514074e-05,
+      "loss": 3.9778,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2764781379003554e-05,
+      "loss": 3.9764,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2756395431493034e-05,
+      "loss": 3.9804,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2748009483982514e-05,
+      "loss": 3.9848,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2739623536471994e-05,
+      "loss": 3.9719,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2731237588961474e-05,
+      "loss": 3.9765,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2722851641450954e-05,
+      "loss": 3.9721,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2714465693940433e-05,
+      "loss": 3.9805,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2706079746429913e-05,
+      "loss": 3.9752,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269769379891939e-05,
+      "loss": 3.9698,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26893242302126e-05,
+      "loss": 3.9762,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268093828270208e-05,
+      "loss": 3.9707,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267255233519156e-05,
+      "loss": 3.987,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266416638768104e-05,
+      "loss": 3.9804,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265578044017052e-05,
+      "loss": 3.9842,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264739449266e-05,
+      "loss": 3.9707,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263900854514948e-05,
+      "loss": 3.9833,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263062259763897e-05,
+      "loss": 3.9867,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262223665012845e-05,
+      "loss": 3.9791,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261386708142166e-05,
+      "loss": 3.9785,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260548113391114e-05,
+      "loss": 3.9716,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259711156520435e-05,
+      "loss": 3.9917,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258872561769383e-05,
+      "loss": 3.9881,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258033967018331e-05,
+      "loss": 3.98,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257195372267279e-05,
+      "loss": 3.9742,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2563584153965996e-05,
+      "loss": 3.9897,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2555198206455476e-05,
+      "loss": 3.9752,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2546812258944956e-05,
+      "loss": 3.9683,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2538426311434436e-05,
+      "loss": 3.9886,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253005674272765e-05,
+      "loss": 3.9765,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252167079521713e-05,
+      "loss": 3.9928,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.251328484770661e-05,
+      "loss": 3.9716,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.03020715713501,
+      "eval_runtime": 307.102,
+      "eval_samples_per_second": 1242.555,
+      "eval_steps_per_second": 38.831,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.250489890019609e-05,
+      "loss": 3.9709,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249651295268557e-05,
+      "loss": 3.9612,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248814338397878e-05,
+      "loss": 3.9719,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247975743646826e-05,
+      "loss": 3.986,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247137148895774e-05,
+      "loss": 3.9812,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246298554144722e-05,
+      "loss": 3.9779,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24545995939367e-05,
+      "loss": 3.9752,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2446213646426174e-05,
+      "loss": 3.9709,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243782769891566e-05,
+      "loss": 3.9707,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2429458130208876e-05,
+      "loss": 3.9762,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2421072182698356e-05,
+      "loss": 3.9821,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.241268623518783e-05,
+      "loss": 3.9837,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.240430028767731e-05,
+      "loss": 3.987,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239591434016679e-05,
+      "loss": 3.9645,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238754477146e-05,
+      "loss": 3.9784,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2379175202753214e-05,
+      "loss": 3.9556,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2370789255242694e-05,
+      "loss": 3.9727,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2362403307732174e-05,
+      "loss": 3.9726,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235401736022165e-05,
+      "loss": 3.9727,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.234563141271113e-05,
+      "loss": 3.966,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2337245465200614e-05,
+      "loss": 3.9903,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2328859517690094e-05,
+      "loss": 3.9774,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2320473570179574e-05,
+      "loss": 3.9823,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2312087622669054e-05,
+      "loss": 3.9768,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2303701675158534e-05,
+      "loss": 3.9833,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2295315727648014e-05,
+      "loss": 3.9668,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2286929780137494e-05,
+      "loss": 3.9692,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.22785602114307e-05,
+      "loss": 3.9809,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227019064272391e-05,
+      "loss": 3.9718,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226180469521339e-05,
+      "loss": 3.9592,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.225341874770287e-05,
+      "loss": 3.971,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.224503280019235e-05,
+      "loss": 3.9739,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.223666323148556e-05,
+      "loss": 3.9843,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222827728397505e-05,
+      "loss": 3.9774,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221989133646453e-05,
+      "loss": 3.9732,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221152176775774e-05,
+      "loss": 3.9835,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.220313582024722e-05,
+      "loss": 3.9833,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.21947498727367e-05,
+      "loss": 3.9584,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2186363925226177e-05,
+      "loss": 3.9706,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2177977977715657e-05,
+      "loss": 3.9606,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2169592030205137e-05,
+      "loss": 3.9667,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2161206082694616e-05,
+      "loss": 3.9801,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2152820135184096e-05,
+      "loss": 3.9808,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2144434187673576e-05,
+      "loss": 3.9697,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2136048240163056e-05,
+      "loss": 3.9769,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2127662292652536e-05,
+      "loss": 3.9739,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2119276345142016e-05,
+      "loss": 3.97,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211090677643523e-05,
+      "loss": 3.9801,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210253720772844e-05,
+      "loss": 3.9517,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209415126021792e-05,
+      "loss": 3.979,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20857653127074e-05,
+      "loss": 3.9737,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207737936519688e-05,
+      "loss": 3.9812,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206899341768636e-05,
+      "loss": 3.9653,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2060607470175834e-05,
+      "loss": 3.966,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2052221522665314e-05,
+      "loss": 3.9617,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2043835575154794e-05,
+      "loss": 3.9618,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203546600644801e-05,
+      "loss": 3.9732,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202708005893748e-05,
+      "loss": 3.9565,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201869411142697e-05,
+      "loss": 3.9816,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201030816391645e-05,
+      "loss": 3.9702,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.200193859520966e-05,
+      "loss": 3.9589,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.199355264769914e-05,
+      "loss": 3.9744,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.198516670018862e-05,
+      "loss": 3.9565,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1976797131481835e-05,
+      "loss": 3.9628,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196841118397131e-05,
+      "loss": 3.9676,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196002523646079e-05,
+      "loss": 3.977,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.195163928895027e-05,
+      "loss": 3.9558,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.194325334143975e-05,
+      "loss": 3.9666,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.193486739392923e-05,
+      "loss": 3.953,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.192648144641871e-05,
+      "loss": 3.9656,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191809549890819e-05,
+      "loss": 3.9759,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1909725930201404e-05,
+      "loss": 3.9738,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1901339982690883e-05,
+      "loss": 3.971,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1892954035180363e-05,
+      "loss": 3.9717,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1884568087669843e-05,
+      "loss": 3.9841,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187619851896305e-05,
+      "loss": 3.9639,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186782895025626e-05,
+      "loss": 3.9712,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185944300274574e-05,
+      "loss": 3.9598,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185105705523522e-05,
+      "loss": 3.9628,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18426711077247e-05,
+      "loss": 3.9712,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.183428516021418e-05,
+      "loss": 3.9629,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182589921270366e-05,
+      "loss": 3.9758,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.181751326519314e-05,
+      "loss": 3.98,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.180912731768262e-05,
+      "loss": 3.9667,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.180075774897584e-05,
+      "loss": 3.9548,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.179237180146532e-05,
+      "loss": 3.9624,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.17839858539548e-05,
+      "loss": 3.9723,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.177559990644428e-05,
+      "loss": 3.9666,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1767230337737486e-05,
+      "loss": 3.9728,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1758860769030695e-05,
+      "loss": 3.9733,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1750474821520175e-05,
+      "loss": 3.965,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1742088874009655e-05,
+      "loss": 3.9615,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1733702926499135e-05,
+      "loss": 3.9632,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1725333357792344e-05,
+      "loss": 3.9695,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.171694741028183e-05,
+      "loss": 3.9746,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170856146277131e-05,
+      "loss": 3.9685,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170017551526079e-05,
+      "loss": 3.9626,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169178956775027e-05,
+      "loss": 3.9576,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168341999904348e-05,
+      "loss": 3.9778,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.167503405153296e-05,
+      "loss": 3.9729,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.166664810402244e-05,
+      "loss": 3.9685,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.165826215651192e-05,
+      "loss": 3.9663,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16498762090014e-05,
+      "loss": 3.9589,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164149026149088e-05,
+      "loss": 3.9657,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.163310431398036e-05,
+      "loss": 3.9672,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.162471836646984e-05,
+      "loss": 3.955,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161634879776305e-05,
+      "loss": 3.9614,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1607979229056265e-05,
+      "loss": 3.9636,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1599593281545744e-05,
+      "loss": 3.9744,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1591207334035224e-05,
+      "loss": 3.956,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1582821386524704e-05,
+      "loss": 3.966,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1574435439014184e-05,
+      "loss": 3.9568,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1566049491503664e-05,
+      "loss": 3.9753,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1557679922796873e-05,
+      "loss": 3.9597,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154929397528635e-05,
+      "loss": 3.9655,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154090802777583e-05,
+      "loss": 3.9706,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153252208026531e-05,
+      "loss": 3.9796,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152415251155852e-05,
+      "loss": 3.9651,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1515766564048e-05,
+      "loss": 3.9598,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150738061653748e-05,
+      "loss": 3.9728,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149899466902696e-05,
+      "loss": 3.9726,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149060872151644e-05,
+      "loss": 3.9537,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.148222277400592e-05,
+      "loss": 3.9646,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14738368264954e-05,
+      "loss": 3.9624,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.146545087898488e-05,
+      "loss": 3.9697,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.145706493147436e-05,
+      "loss": 3.9647,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144867898396384e-05,
+      "loss": 3.9538,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144029303645332e-05,
+      "loss": 3.9675,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.143192346774653e-05,
+      "loss": 3.9591,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.142353752023601e-05,
+      "loss": 3.9678,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.141515157272549e-05,
+      "loss": 3.9716,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140676562521497e-05,
+      "loss": 3.9669,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139837967770445e-05,
+      "loss": 3.9644,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138999373019393e-05,
+      "loss": 3.9647,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138160778268342e-05,
+      "loss": 3.9783,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13732218351729e-05,
+      "loss": 3.9664,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1364852266466107e-05,
+      "loss": 3.9674,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1356482697759316e-05,
+      "loss": 3.9561,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1348096750248796e-05,
+      "loss": 3.9826,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1339710802738276e-05,
+      "loss": 3.9712,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1331324855227756e-05,
+      "loss": 3.9713,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1322938907717236e-05,
+      "loss": 3.9608,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1314569339010445e-05,
+      "loss": 3.9742,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1306183391499925e-05,
+      "loss": 3.9632,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1297797443989405e-05,
+      "loss": 3.9533,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1289411496478884e-05,
+      "loss": 3.9795,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128102554896837e-05,
+      "loss": 3.9651,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127263960145785e-05,
+      "loss": 3.979,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126427003275106e-05,
+      "loss": 3.9585,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0234694480896,
+      "eval_runtime": 307.091,
+      "eval_samples_per_second": 1242.599,
+      "eval_steps_per_second": 38.832,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125588408524054e-05,
+      "loss": 3.9262,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124749813773002e-05,
+      "loss": 3.9501,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.12391121902195e-05,
+      "loss": 3.9588,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123072624270898e-05,
+      "loss": 3.9722,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122234029519845e-05,
+      "loss": 3.9676,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121395434768793e-05,
+      "loss": 3.9694,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120556840017741e-05,
+      "loss": 3.9633,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119718245266689e-05,
+      "loss": 3.9594,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118879650515637e-05,
+      "loss": 3.9583,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118041055764585e-05,
+      "loss": 3.9638,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117202461013533e-05,
+      "loss": 3.9683,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116363866262482e-05,
+      "loss": 3.9721,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.115526909391803e-05,
+      "loss": 3.9745,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114688314640751e-05,
+      "loss": 3.9543,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113849719889699e-05,
+      "loss": 3.9648,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113011125138647e-05,
+      "loss": 3.9431,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112172530387595e-05,
+      "loss": 3.9607,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111333935636543e-05,
+      "loss": 3.9606,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110495340885491e-05,
+      "loss": 3.9598,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109656746134439e-05,
+      "loss": 3.9569,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108818151383387e-05,
+      "loss": 3.9737,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107981194512708e-05,
+      "loss": 3.9679,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107142599761656e-05,
+      "loss": 3.9653,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.106304005010604e-05,
+      "loss": 3.9648,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.105465410259552e-05,
+      "loss": 3.9734,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1046268155085e-05,
+      "loss": 3.9551,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103788220757448e-05,
+      "loss": 3.9574,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102949626006396e-05,
+      "loss": 3.9711,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102111031255344e-05,
+      "loss": 3.9571,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1012740743846646e-05,
+      "loss": 3.9456,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1004354796336126e-05,
+      "loss": 3.9606,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0995968848825606e-05,
+      "loss": 3.9632,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0987582901315086e-05,
+      "loss": 3.9703,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0979196953804566e-05,
+      "loss": 3.9638,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0970827385097775e-05,
+      "loss": 3.9635,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.096245781639099e-05,
+      "loss": 3.9715,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.095407186888047e-05,
+      "loss": 3.9683,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094568592136995e-05,
+      "loss": 3.9482,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093729997385943e-05,
+      "loss": 3.9637,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092893040515264e-05,
+      "loss": 3.9467,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092054445764212e-05,
+      "loss": 3.9578,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09121585101316e-05,
+      "loss": 3.9613,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.090377256262108e-05,
+      "loss": 3.9692,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089538661511056e-05,
+      "loss": 3.9611,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088700066760004e-05,
+      "loss": 3.9614,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087861472008952e-05,
+      "loss": 3.9607,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087024515138273e-05,
+      "loss": 3.9593,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086185920387221e-05,
+      "loss": 3.9702,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0853473256361696e-05,
+      "loss": 3.9399,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0845087308851176e-05,
+      "loss": 3.9654,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0836701361340656e-05,
+      "loss": 3.9638,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0828315413830135e-05,
+      "loss": 3.9693,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0819945845123345e-05,
+      "loss": 3.9547,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0811559897612825e-05,
+      "loss": 3.956,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0803173950102304e-05,
+      "loss": 3.9443,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0794788002591784e-05,
+      "loss": 3.948,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0786402055081264e-05,
+      "loss": 3.9684,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0778016107570744e-05,
+      "loss": 3.9459,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0769630160060224e-05,
+      "loss": 3.9649,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07612442125497e-05,
+      "loss": 3.9609,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075287464384291e-05,
+      "loss": 3.9453,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074448869633239e-05,
+      "loss": 3.9615,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073610274882188e-05,
+      "loss": 3.9455,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072771680131135e-05,
+      "loss": 3.9495,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071934723260457e-05,
+      "loss": 3.9535,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071096128509405e-05,
+      "loss": 3.9668,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070257533758353e-05,
+      "loss": 3.9431,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0694189390073e-05,
+      "loss": 3.9503,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.068580344256248e-05,
+      "loss": 3.9406,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067741749505196e-05,
+      "loss": 3.9547,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066904792634517e-05,
+      "loss": 3.9651,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066066197883465e-05,
+      "loss": 3.964,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.065227603132413e-05,
+      "loss": 3.9537,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.064389008381362e-05,
+      "loss": 3.9621,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06355041363031e-05,
+      "loss": 3.9708,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.062713456759631e-05,
+      "loss": 3.9538,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061874862008579e-05,
+      "loss": 3.9562,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061036267257527e-05,
+      "loss": 3.9514,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060197672506475e-05,
+      "loss": 3.95,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059359077755423e-05,
+      "loss": 3.9594,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058520483004371e-05,
+      "loss": 3.9507,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057681888253319e-05,
+      "loss": 3.9656,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0568432935022667e-05,
+      "loss": 3.9663,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0560063366315876e-05,
+      "loss": 3.9569,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0551677418805356e-05,
+      "loss": 3.9418,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0543291471294836e-05,
+      "loss": 3.9509,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0534905523784316e-05,
+      "loss": 3.9594,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052653595507753e-05,
+      "loss": 3.9561,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051815000756701e-05,
+      "loss": 3.9589,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.050978043886022e-05,
+      "loss": 3.9637,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.05013944913497e-05,
+      "loss": 3.9545,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.049300854383918e-05,
+      "loss": 3.948,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.048462259632866e-05,
+      "loss": 3.9478,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.047623664881814e-05,
+      "loss": 3.9596,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.046786708011135e-05,
+      "loss": 3.959,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0459481132600833e-05,
+      "loss": 3.9602,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0451095185090313e-05,
+      "loss": 3.9486,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442709237579793e-05,
+      "loss": 3.9488,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0434323290069273e-05,
+      "loss": 3.9627,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0425937342558753e-05,
+      "loss": 3.9613,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041756777385196e-05,
+      "loss": 3.9614,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040918182634144e-05,
+      "loss": 3.9483,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0400795878830925e-05,
+      "loss": 3.9482,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0392409931320405e-05,
+      "loss": 3.9569,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0384023983809885e-05,
+      "loss": 3.9547,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375638036299358e-05,
+      "loss": 3.9411,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0367252088788838e-05,
+      "loss": 3.9451,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0358866141278318e-05,
+      "loss": 3.9547,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0350496572571534e-05,
+      "loss": 3.9578,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.034211062506101e-05,
+      "loss": 3.9489,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033372467755049e-05,
+      "loss": 3.9587,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032533873003997e-05,
+      "loss": 3.9385,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0316969161333186e-05,
+      "loss": 3.966,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0308583213822663e-05,
+      "loss": 3.9439,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0300197266312143e-05,
+      "loss": 3.9545,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0291811318801623e-05,
+      "loss": 3.9603,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0283425371291103e-05,
+      "loss": 3.9651,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0275072181388048e-05,
+      "loss": 3.9529,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0266686233877528e-05,
+      "loss": 3.9486,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0258300286367008e-05,
+      "loss": 3.9575,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0249914338856484e-05,
+      "loss": 3.961,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0241528391345964e-05,
+      "loss": 3.9462,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0233142443835444e-05,
+      "loss": 3.9496,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0224756496324924e-05,
+      "loss": 3.9497,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0216370548814404e-05,
+      "loss": 3.9619,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0208000980107613e-05,
+      "loss": 3.9471,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0199615032597096e-05,
+      "loss": 3.9453,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0191229085086576e-05,
+      "loss": 3.9491,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182843137576056e-05,
+      "loss": 3.9537,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174457190065536e-05,
+      "loss": 3.9462,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0166087621358745e-05,
+      "loss": 3.9641,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0157701673848225e-05,
+      "loss": 3.9597,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0149332105141438e-05,
+      "loss": 3.9471,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0140946157630918e-05,
+      "loss": 3.9512,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0132560210120398e-05,
+      "loss": 3.9633,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0124174262609878e-05,
+      "loss": 3.9558,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0115788315099358e-05,
+      "loss": 3.9564,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0107402367588838e-05,
+      "loss": 3.9501,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0099016420078318e-05,
+      "loss": 3.9674,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0090630472567797e-05,
+      "loss": 3.958,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.008226090386101e-05,
+      "loss": 3.9608,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.007387495635049e-05,
+      "loss": 3.9465,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.00655053876437e-05,
+      "loss": 3.9656,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.005711944013318e-05,
+      "loss": 3.9517,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004873349262266e-05,
+      "loss": 3.9419,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0040347545112142e-05,
+      "loss": 3.9663,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0031961597601622e-05,
+      "loss": 3.9541,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023575650091102e-05,
+      "loss": 3.9626,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0015189702580582e-05,
+      "loss": 3.9497,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0006803755070062e-05,
+      "loss": 3.946,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.016993045806885,
+      "eval_runtime": 303.4965,
+      "eval_samples_per_second": 1257.316,
+      "eval_steps_per_second": 39.292,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.999843418636327e-05,
+      "loss": 3.9399,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.999004823885275e-05,
+      "loss": 3.9442,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9981662291342234e-05,
+      "loss": 3.9623,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9973276343831714e-05,
+      "loss": 3.9543,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9964906775124924e-05,
+      "loss": 3.9592,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9956520827614403e-05,
+      "loss": 3.9513,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9948134880103883e-05,
+      "loss": 3.9469,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9939748932593363e-05,
+      "loss": 3.9423,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9931379363886576e-05,
+      "loss": 3.9573,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9922993416376056e-05,
+      "loss": 3.9505,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9914607468865536e-05,
+      "loss": 3.967,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9906237900158745e-05,
+      "loss": 3.9643,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9897851952648225e-05,
+      "loss": 3.943,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9889466005137705e-05,
+      "loss": 3.9469,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9881080057627188e-05,
+      "loss": 3.9338,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9872694110116668e-05,
+      "loss": 3.9425,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.986430816260614e-05,
+      "loss": 3.9527,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.985592221509562e-05,
+      "loss": 3.9471,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.98475362675851e-05,
+      "loss": 3.9448,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.983915032007458e-05,
+      "loss": 3.9626,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9830780751367794e-05,
+      "loss": 3.9525,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9822394803857274e-05,
+      "loss": 3.9546,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9814008856346754e-05,
+      "loss": 3.956,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9805622908836233e-05,
+      "loss": 3.9651,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9797253340129443e-05,
+      "loss": 3.9407,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.978888377142266e-05,
+      "loss": 3.946,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9780497823912142e-05,
+      "loss": 3.9583,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9772111876401615e-05,
+      "loss": 3.9421,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9763725928891095e-05,
+      "loss": 3.9387,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9755339981380575e-05,
+      "loss": 3.9446,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9746954033870055e-05,
+      "loss": 3.9519,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9738568086359535e-05,
+      "loss": 3.9591,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9730182138849018e-05,
+      "loss": 3.9513,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9721812570142227e-05,
+      "loss": 3.9521,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9713443001435436e-05,
+      "loss": 3.9591,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9705057053924916e-05,
+      "loss": 3.9535,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9696671106414396e-05,
+      "loss": 3.9401,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.968828515890388e-05,
+      "loss": 3.9506,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967989921139336e-05,
+      "loss": 3.9327,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967151326388284e-05,
+      "loss": 3.9495,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.966312731637232e-05,
+      "loss": 3.9488,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.965475774766553e-05,
+      "loss": 3.9577,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.964637180015501e-05,
+      "loss": 3.9493,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.963798585264449e-05,
+      "loss": 3.951,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962959990513397e-05,
+      "loss": 3.9492,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962123033642718e-05,
+      "loss": 3.948,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.961284438891666e-05,
+      "loss": 3.9591,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.960445844140614e-05,
+      "loss": 3.9271,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.959607249389562e-05,
+      "loss": 3.9526,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.95876865463851e-05,
+      "loss": 3.9545,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957930059887458e-05,
+      "loss": 3.9564,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957091465136406e-05,
+      "loss": 3.9431,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9562528703853544e-05,
+      "loss": 3.9441,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9554159135146753e-05,
+      "loss": 3.9353,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9545773187636233e-05,
+      "loss": 3.9359,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9537387240125713e-05,
+      "loss": 3.9587,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9529017671418922e-05,
+      "loss": 3.9332,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9520631723908405e-05,
+      "loss": 3.9528,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9512245776397885e-05,
+      "loss": 3.9494,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9503859828887365e-05,
+      "loss": 3.9327,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9495473881376845e-05,
+      "loss": 3.9494,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9487087933866325e-05,
+      "loss": 3.9368,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.94787019863558e-05,
+      "loss": 3.9399,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9470316038845282e-05,
+      "loss": 3.9408,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9461930091334762e-05,
+      "loss": 3.9577,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9453560522627978e-05,
+      "loss": 3.9335,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.944517457511745e-05,
+      "loss": 3.9379,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.943678862760693e-05,
+      "loss": 3.9298,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.942840268009641e-05,
+      "loss": 3.9421,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9420033111389627e-05,
+      "loss": 3.9503,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9411647163879103e-05,
+      "loss": 3.9539,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9403261216368583e-05,
+      "loss": 3.9422,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9394875268858063e-05,
+      "loss": 3.9543,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9386489321347543e-05,
+      "loss": 3.959,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9378119752640752e-05,
+      "loss": 3.9443,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9369733805130235e-05,
+      "loss": 3.9457,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.936136423642345e-05,
+      "loss": 3.9399,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9352978288912925e-05,
+      "loss": 3.9409,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9344592341402404e-05,
+      "loss": 3.9488,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9336206393891884e-05,
+      "loss": 3.935,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9327820446381364e-05,
+      "loss": 3.958,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9319434498870844e-05,
+      "loss": 3.9575,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9311048551360328e-05,
+      "loss": 3.9421,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9302662603849808e-05,
+      "loss": 3.9313,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9294293035143017e-05,
+      "loss": 3.9395,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9285907087632497e-05,
+      "loss": 3.9473,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9277521140121977e-05,
+      "loss": 3.9507,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9269135192611457e-05,
+      "loss": 3.9449,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9260749245100937e-05,
+      "loss": 3.9535,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9252396055197878e-05,
+      "loss": 3.9455,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9244010107687358e-05,
+      "loss": 3.9338,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9235624160176838e-05,
+      "loss": 3.9356,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9227238212666318e-05,
+      "loss": 3.9499,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9218852265155798e-05,
+      "loss": 3.9451,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.921046631764528e-05,
+      "loss": 3.9528,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.920208037013476e-05,
+      "loss": 3.9369,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.919369442262424e-05,
+      "loss": 3.9389,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.918532485391745e-05,
+      "loss": 3.9538,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.917693890640693e-05,
+      "loss": 3.9504,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.916855295889641e-05,
+      "loss": 3.9472,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.916016701138589e-05,
+      "loss": 3.9436,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9151781063875374e-05,
+      "loss": 3.9376,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9143395116364854e-05,
+      "loss": 3.9409,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9135009168854333e-05,
+      "loss": 3.9486,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9126623221343813e-05,
+      "loss": 3.9273,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9118253652637023e-05,
+      "loss": 3.9376,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9109867705126502e-05,
+      "loss": 3.9467,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9101481757615982e-05,
+      "loss": 3.9479,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.909309581010546e-05,
+      "loss": 3.9334,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9084726241398675e-05,
+      "loss": 3.9491,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9076340293888155e-05,
+      "loss": 3.9314,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9067954346377635e-05,
+      "loss": 3.9582,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.905956839886711e-05,
+      "loss": 3.9298,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9051215208964056e-05,
+      "loss": 3.945,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9042829261453536e-05,
+      "loss": 3.9461,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9034443313943016e-05,
+      "loss": 3.9571,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9026073745236225e-05,
+      "loss": 3.9434,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9017687797725705e-05,
+      "loss": 3.9409,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9009301850215185e-05,
+      "loss": 3.9449,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.900091590270467e-05,
+      "loss": 3.9477,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.899252995519415e-05,
+      "loss": 3.9382,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.898414400768363e-05,
+      "loss": 3.9414,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.897575806017311e-05,
+      "loss": 3.9356,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.896737211266258e-05,
+      "loss": 3.9563,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8959002543955798e-05,
+      "loss": 3.9322,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8950616596445278e-05,
+      "loss": 3.9388,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.894223064893476e-05,
+      "loss": 3.9367,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8933844701424234e-05,
+      "loss": 3.9436,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8925458753913714e-05,
+      "loss": 3.9377,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8917072806403194e-05,
+      "loss": 3.9572,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8908703237696406e-05,
+      "loss": 3.9476,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8900317290185886e-05,
+      "loss": 3.9435,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8891931342675366e-05,
+      "loss": 3.9383,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8883545395164846e-05,
+      "loss": 3.9493,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8875159447654326e-05,
+      "loss": 3.9481,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8866773500143806e-05,
+      "loss": 3.9483,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8858387552633286e-05,
+      "loss": 3.9386,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8850001605122766e-05,
+      "loss": 3.9574,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.884163203641598e-05,
+      "loss": 3.9468,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.883324608890546e-05,
+      "loss": 3.9537,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.882486014139494e-05,
+      "loss": 3.9358,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.881647419388442e-05,
+      "loss": 3.9568,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.880812100398136e-05,
+      "loss": 3.945,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.879973505647084e-05,
+      "loss": 3.9271,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.879134910896032e-05,
+      "loss": 3.9586,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.87829631614498e-05,
+      "loss": 3.9462,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.877457721393928e-05,
+      "loss": 3.9477,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.876619126642876e-05,
+      "loss": 3.9432,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.875780531891824e-05,
+      "loss": 3.9354,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.012063503265381,
+      "eval_runtime": 292.5959,
+      "eval_samples_per_second": 1304.157,
+      "eval_steps_per_second": 40.756,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.874941937140772e-05,
+      "loss": 3.9271,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.87410334238972e-05,
+      "loss": 3.9369,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8732647476386683e-05,
+      "loss": 3.9521,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8724261528876163e-05,
+      "loss": 3.9425,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8715875581365643e-05,
+      "loss": 3.9498,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8707489633855123e-05,
+      "loss": 3.9439,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8699103686344596e-05,
+      "loss": 3.936,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8690717738834076e-05,
+      "loss": 3.9344,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.868233179132356e-05,
+      "loss": 3.9443,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.867394584381304e-05,
+      "loss": 3.9431,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.866555989630252e-05,
+      "loss": 3.9551,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8657173948792e-05,
+      "loss": 3.9516,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.864880438008521e-05,
+      "loss": 3.9399,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.864041843257469e-05,
+      "loss": 3.9336,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8632032485064168e-05,
+      "loss": 3.9284,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.862364653755365e-05,
+      "loss": 3.9311,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.861526059004313e-05,
+      "loss": 3.9419,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.860689102133634e-05,
+      "loss": 3.9383,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.859850507382582e-05,
+      "loss": 3.9341,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.85901191263153e-05,
+      "loss": 3.9566,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858173317880478e-05,
+      "loss": 3.9434,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.857334723129426e-05,
+      "loss": 3.9439,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.856496128378374e-05,
+      "loss": 3.9442,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8556575336273224e-05,
+      "loss": 3.9582,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8548189388762704e-05,
+      "loss": 3.9294,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8539803441252184e-05,
+      "loss": 3.938,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8531417493741657e-05,
+      "loss": 3.9469,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8523047925034873e-05,
+      "loss": 3.9371,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8514661977524353e-05,
+      "loss": 3.9281,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8506276030013833e-05,
+      "loss": 3.9317,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.849789008250331e-05,
+      "loss": 3.9456,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.848950413499279e-05,
+      "loss": 3.9448,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8481134566286005e-05,
+      "loss": 3.9415,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8472748618775478e-05,
+      "loss": 3.9438,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8464379050068694e-05,
+      "loss": 3.948,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8455993102558177e-05,
+      "loss": 3.9437,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8447607155047657e-05,
+      "loss": 3.9322,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843922120753713e-05,
+      "loss": 3.9396,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843083526002661e-05,
+      "loss": 3.926,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.842244931251609e-05,
+      "loss": 3.9389,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.841406336500557e-05,
+      "loss": 3.9385,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8405677417495054e-05,
+      "loss": 3.9501,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8397291469984534e-05,
+      "loss": 3.9396,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8388905522474014e-05,
+      "loss": 3.9431,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8380519574963494e-05,
+      "loss": 3.9383,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8372133627452974e-05,
+      "loss": 3.9413,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8363764058746183e-05,
+      "loss": 3.9506,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8355378111235663e-05,
+      "loss": 3.9195,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8346992163725146e-05,
+      "loss": 3.936,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8338606216214626e-05,
+      "loss": 3.9502,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8330236647507835e-05,
+      "loss": 3.9455,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8321850699997315e-05,
+      "loss": 3.9339,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8313464752486795e-05,
+      "loss": 3.9311,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8305078804976275e-05,
+      "loss": 3.9331,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8296709236269487e-05,
+      "loss": 3.9219,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8288323288758967e-05,
+      "loss": 3.9497,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8279937341248447e-05,
+      "loss": 3.9261,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8271551393737927e-05,
+      "loss": 3.9443,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8263181825031136e-05,
+      "loss": 3.9378,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8254795877520616e-05,
+      "loss": 3.927,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.82464099300101e-05,
+      "loss": 3.9353,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.823802398249958e-05,
+      "loss": 3.9295,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.822965441379279e-05,
+      "loss": 3.9275,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.822126846628227e-05,
+      "loss": 3.9334,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.821288251877175e-05,
+      "loss": 3.952,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.820449657126123e-05,
+      "loss": 3.9247,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.819611062375071e-05,
+      "loss": 3.9281,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8187724676240192e-05,
+      "loss": 3.9181,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8179338728729672e-05,
+      "loss": 3.9314,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8170952781219145e-05,
+      "loss": 3.9467,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8162566833708625e-05,
+      "loss": 3.9422,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.815419726500184e-05,
+      "loss": 3.9346,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8145811317491317e-05,
+      "loss": 3.9448,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8137441748784533e-05,
+      "loss": 3.9496,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8129055801274013e-05,
+      "loss": 3.9341,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8120669853763493e-05,
+      "loss": 3.9324,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8112283906252966e-05,
+      "loss": 3.9336,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8103897958742446e-05,
+      "loss": 3.9319,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.809551201123193e-05,
+      "loss": 3.9372,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.808712606372141e-05,
+      "loss": 3.9288,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.807875649501462e-05,
+      "loss": 3.9443,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.80703705475041e-05,
+      "loss": 3.9431,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.806198459999358e-05,
+      "loss": 3.9381,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.805359865248306e-05,
+      "loss": 3.921,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.804521270497254e-05,
+      "loss": 3.929,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8036826757462022e-05,
+      "loss": 3.9379,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8028440809951502e-05,
+      "loss": 3.9399,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8020054862440982e-05,
+      "loss": 3.939,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8011668914930462e-05,
+      "loss": 3.9385,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.80033157250274e-05,
+      "loss": 3.938,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7994929777516883e-05,
+      "loss": 3.9285,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7986543830006363e-05,
+      "loss": 3.9245,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7978157882495843e-05,
+      "loss": 3.941,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7969771934985323e-05,
+      "loss": 3.9346,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7961402366278532e-05,
+      "loss": 3.9406,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7953016418768012e-05,
+      "loss": 3.9333,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7944630471257492e-05,
+      "loss": 3.926,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7936244523746972e-05,
+      "loss": 3.9428,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7927858576236456e-05,
+      "loss": 3.9446,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7919472628725936e-05,
+      "loss": 3.9381,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7911086681215415e-05,
+      "loss": 3.9343,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7902717112508625e-05,
+      "loss": 3.925,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7894331164998105e-05,
+      "loss": 3.9341,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7885945217487584e-05,
+      "loss": 3.9401,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7877559269977064e-05,
+      "loss": 3.9196,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7869189701270277e-05,
+      "loss": 3.9263,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7860803753759757e-05,
+      "loss": 3.9334,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7852417806249237e-05,
+      "loss": 3.9418,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7844031858738717e-05,
+      "loss": 3.9234,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7835645911228197e-05,
+      "loss": 3.9365,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.782727634252141e-05,
+      "loss": 3.9252,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.781889039501089e-05,
+      "loss": 3.9449,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.781050444750037e-05,
+      "loss": 3.9223,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.780211849998985e-05,
+      "loss": 3.934,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7793748931283058e-05,
+      "loss": 3.941,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7785362983772538e-05,
+      "loss": 3.9425,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.777699341506575e-05,
+      "loss": 3.9361,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.776860746755523e-05,
+      "loss": 3.9285,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.776022152004471e-05,
+      "loss": 3.9352,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.775183557253419e-05,
+      "loss": 3.9415,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.774344962502367e-05,
+      "loss": 3.9257,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.773506367751315e-05,
+      "loss": 3.9322,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7726677730002627e-05,
+      "loss": 3.9269,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7718308161295843e-05,
+      "loss": 3.9449,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7709922213785323e-05,
+      "loss": 3.92,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7701536266274803e-05,
+      "loss": 3.9327,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7693150318764276e-05,
+      "loss": 3.9262,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7684764371253756e-05,
+      "loss": 3.9325,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.767637842374324e-05,
+      "loss": 3.9269,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.766799247623272e-05,
+      "loss": 3.947,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76596065287222e-05,
+      "loss": 3.9414,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7651236960015408e-05,
+      "loss": 3.9335,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7642851012504888e-05,
+      "loss": 3.9267,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7634465064994368e-05,
+      "loss": 3.9397,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7626079117483848e-05,
+      "loss": 3.937,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.761769316997333e-05,
+      "loss": 3.9426,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.760930722246281e-05,
+      "loss": 3.9235,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.760092127495229e-05,
+      "loss": 3.9483,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.759253532744177e-05,
+      "loss": 3.9363,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.758416575873498e-05,
+      "loss": 3.9464,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.757577981122446e-05,
+      "loss": 3.929,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7567410242517673e-05,
+      "loss": 3.9451,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7559024295007153e-05,
+      "loss": 3.9357,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7550638347496633e-05,
+      "loss": 3.9235,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7542252399986113e-05,
+      "loss": 3.9439,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7533866452475593e-05,
+      "loss": 3.94,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7525480504965073e-05,
+      "loss": 3.937,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7517094557454553e-05,
+      "loss": 3.9321,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7508708609944033e-05,
+      "loss": 3.9277,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.007241725921631,
+      "eval_runtime": 295.6522,
+      "eval_samples_per_second": 1290.675,
+      "eval_steps_per_second": 40.335,
+      "step": 1373760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7500339041237245e-05,
+      "loss": 3.9209,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7491953093726725e-05,
+      "loss": 3.9253,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7483567146216205e-05,
+      "loss": 3.9424,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7475181198705685e-05,
+      "loss": 3.9328,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7466811629998894e-05,
+      "loss": 3.9416,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7458425682488377e-05,
+      "loss": 3.928,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7450039734977857e-05,
+      "loss": 3.9285,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7441653787467337e-05,
+      "loss": 3.9241,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7433284218760546e-05,
+      "loss": 3.9332,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7424898271250026e-05,
+      "loss": 3.936,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7416512323739506e-05,
+      "loss": 3.9407,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7408126376228986e-05,
+      "loss": 3.9419,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.73997568075222e-05,
+      "loss": 3.9281,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739137086001168e-05,
+      "loss": 3.9248,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.738298491250116e-05,
+      "loss": 3.9262,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7374598964990632e-05,
+      "loss": 3.9174,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7366213017480115e-05,
+      "loss": 3.9317,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7357827069969595e-05,
+      "loss": 3.9263,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7349441122459075e-05,
+      "loss": 3.9259,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7341071553752284e-05,
+      "loss": 3.95,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7332685606241764e-05,
+      "loss": 3.9326,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7324299658731244e-05,
+      "loss": 3.9352,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7315913711220724e-05,
+      "loss": 3.9361,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7307527763710204e-05,
+      "loss": 3.9469,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7299141816199687e-05,
+      "loss": 3.9154,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7290772247492896e-05,
+      "loss": 3.9287,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7282386299982376e-05,
+      "loss": 3.9354,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7274000352471856e-05,
+      "loss": 3.9318,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7265614404961336e-05,
+      "loss": 3.9212,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7257228457450816e-05,
+      "loss": 3.9189,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7248842509940296e-05,
+      "loss": 3.937,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.724045656242978e-05,
+      "loss": 3.9385,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.723208699372299e-05,
+      "loss": 3.9292,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.722370104621247e-05,
+      "loss": 3.9375,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7215331477505678e-05,
+      "loss": 3.9367,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7206945529995158e-05,
+      "loss": 3.9303,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.719855958248464e-05,
+      "loss": 3.923,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.719017363497412e-05,
+      "loss": 3.9361,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.71817876874636e-05,
+      "loss": 3.9125,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.717340173995308e-05,
+      "loss": 3.9285,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.716501579244256e-05,
+      "loss": 3.9302,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.715662984493204e-05,
+      "loss": 3.9399,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.714824389742152e-05,
+      "loss": 3.9295,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7139857949910997e-05,
+      "loss": 3.9309,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7131488381204213e-05,
+      "loss": 3.9326,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7123102433693693e-05,
+      "loss": 3.9279,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7114716486183173e-05,
+      "loss": 3.9412,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7106330538672646e-05,
+      "loss": 3.9093,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7097944591162126e-05,
+      "loss": 3.9307,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.708955864365161e-05,
+      "loss": 3.9363,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.708117269614109e-05,
+      "loss": 3.9367,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.707278674863057e-05,
+      "loss": 3.9246,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.706440080112005e-05,
+      "loss": 3.9179,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.705603123241326e-05,
+      "loss": 3.9245,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.704764528490274e-05,
+      "loss": 3.9082,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.703925933739222e-05,
+      "loss": 3.9415,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7030873389881702e-05,
+      "loss": 3.9206,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7022487442371182e-05,
+      "loss": 3.9271,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.701411787366439e-05,
+      "loss": 3.9285,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.700573192615387e-05,
+      "loss": 3.9183,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.699734597864335e-05,
+      "loss": 3.9258,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.698896003113283e-05,
+      "loss": 3.9197,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6980590462426043e-05,
+      "loss": 3.92,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6972204514915523e-05,
+      "loss": 3.92,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6963818567405003e-05,
+      "loss": 3.9402,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6955432619894483e-05,
+      "loss": 3.9196,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6947063051187692e-05,
+      "loss": 3.9189,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6938677103677172e-05,
+      "loss": 3.9104,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6930291156166655e-05,
+      "loss": 3.9214,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6921905208656135e-05,
+      "loss": 3.9316,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6913552018753074e-05,
+      "loss": 3.9373,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6905166071242554e-05,
+      "loss": 3.9229,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6896780123732034e-05,
+      "loss": 3.9345,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6888394176221517e-05,
+      "loss": 3.9409,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6880008228710997e-05,
+      "loss": 3.926,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6871638660004206e-05,
+      "loss": 3.9158,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6863252712493686e-05,
+      "loss": 3.9294,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6854866764983166e-05,
+      "loss": 3.9222,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6846480817472646e-05,
+      "loss": 3.9265,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6838094869962126e-05,
+      "loss": 3.9198,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6829725301255338e-05,
+      "loss": 3.9319,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6821339353744818e-05,
+      "loss": 3.9386,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6812953406234298e-05,
+      "loss": 3.9268,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6804567458723778e-05,
+      "loss": 3.9133,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6796181511213258e-05,
+      "loss": 3.9207,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6787795563702738e-05,
+      "loss": 3.9274,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.677942599499595e-05,
+      "loss": 3.9325,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.677104004748543e-05,
+      "loss": 3.9282,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.676265409997491e-05,
+      "loss": 3.9326,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.675426815246439e-05,
+      "loss": 3.9261,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.674588220495387e-05,
+      "loss": 3.9185,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.673751263624708e-05,
+      "loss": 3.9172,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.672912668873656e-05,
+      "loss": 3.9218,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6720740741226043e-05,
+      "loss": 3.9268,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6712354793715523e-05,
+      "loss": 3.9316,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6703968846205003e-05,
+      "loss": 3.9269,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6695599277498212e-05,
+      "loss": 3.9135,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.668721332998769e-05,
+      "loss": 3.9331,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.667882738247717e-05,
+      "loss": 3.9371,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.667044143496665e-05,
+      "loss": 3.9275,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6662055487456128e-05,
+      "loss": 3.9288,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6653685918749344e-05,
+      "loss": 3.9148,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6645299971238824e-05,
+      "loss": 3.9256,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6636914023728304e-05,
+      "loss": 3.9335,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.662852807621778e-05,
+      "loss": 3.9077,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.662014212870726e-05,
+      "loss": 3.9166,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.661175618119674e-05,
+      "loss": 3.9282,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6603386612489956e-05,
+      "loss": 3.9274,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.659500066497943e-05,
+      "loss": 3.9147,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.658661471746891e-05,
+      "loss": 3.9282,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.657822876995839e-05,
+      "loss": 3.914,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6569842822447873e-05,
+      "loss": 3.9336,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6561456874937353e-05,
+      "loss": 3.9175,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6553070927426833e-05,
+      "loss": 3.9203,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6544684979916313e-05,
+      "loss": 3.9342,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.653631541120952e-05,
+      "loss": 3.9358,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6527945842502734e-05,
+      "loss": 3.9279,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6519559894992214e-05,
+      "loss": 3.9226,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6511173947481694e-05,
+      "loss": 3.9235,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6502787999971174e-05,
+      "loss": 3.9276,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6494402052460654e-05,
+      "loss": 3.9217,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6486016104950134e-05,
+      "loss": 3.9253,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6477630157439614e-05,
+      "loss": 3.9162,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6469244209929094e-05,
+      "loss": 3.9315,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6460874641222306e-05,
+      "loss": 3.921,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6452488693711786e-05,
+      "loss": 3.9217,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6444119125004995e-05,
+      "loss": 3.9139,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6435733177494475e-05,
+      "loss": 3.9258,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6427347229983955e-05,
+      "loss": 3.9152,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6418961282473435e-05,
+      "loss": 3.9369,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.641057533496292e-05,
+      "loss": 3.9339,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.64021893874524e-05,
+      "loss": 3.9275,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.639380343994188e-05,
+      "loss": 3.9158,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.638541749243136e-05,
+      "loss": 3.9338,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6377047923724568e-05,
+      "loss": 3.9295,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6368661976214048e-05,
+      "loss": 3.931,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636029240750726e-05,
+      "loss": 3.9187,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.635190645999674e-05,
+      "loss": 3.9325,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.634352051248622e-05,
+      "loss": 3.9317,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.63351345649757e-05,
+      "loss": 3.9343,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.632674861746518e-05,
+      "loss": 3.9224,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.631836266995466e-05,
+      "loss": 3.9305,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6309993101247872e-05,
+      "loss": 3.9263,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6301607153737352e-05,
+      "loss": 3.9158,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6293221206226832e-05,
+      "loss": 3.9327,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6284835258716312e-05,
+      "loss": 3.9349,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6276449311205792e-05,
+      "loss": 3.9258,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6268063363695265e-05,
+      "loss": 3.9258,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.625967741618475e-05,
+      "loss": 3.916,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.002674579620361,
+      "eval_runtime": 296.6972,
+      "eval_samples_per_second": 1286.129,
+      "eval_steps_per_second": 40.192,
+      "step": 1450080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.625129146867423e-05,
+      "loss": 3.9124,
+      "step": 1450496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.624290552116371e-05,
+      "loss": 3.9139,
+      "step": 1451008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.623451957365319e-05,
+      "loss": 3.9339,
+      "step": 1451520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.622613362614267e-05,
+      "loss": 3.9221,
+      "step": 1452032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.621774767863215e-05,
+      "loss": 3.9369,
+      "step": 1452544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.620936173112163e-05,
+      "loss": 3.92,
+      "step": 1453056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.620097578361111e-05,
+      "loss": 3.9233,
+      "step": 1453568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6192589836100588e-05,
+      "loss": 3.9119,
+      "step": 1454080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6184203888590068e-05,
+      "loss": 3.9253,
+      "step": 1454592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.617581794107955e-05,
+      "loss": 3.9242,
+      "step": 1455104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.616743199356903e-05,
+      "loss": 3.9342,
+      "step": 1455616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6159046046058505e-05,
+      "loss": 3.9321,
+      "step": 1456128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6150660098547985e-05,
+      "loss": 3.9191,
+      "step": 1456640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6142274151037465e-05,
+      "loss": 3.9178,
+      "step": 1457152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6133888203526945e-05,
+      "loss": 3.9174,
+      "step": 1457664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6125502256016428e-05,
+      "loss": 3.9123,
+      "step": 1458176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6117116308505908e-05,
+      "loss": 3.917,
+      "step": 1458688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6108730360995388e-05,
+      "loss": 3.9177,
+      "step": 1459200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6100344413484868e-05,
+      "loss": 3.917,
+      "step": 1459712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6091974844778077e-05,
+      "loss": 3.9447,
+      "step": 1460224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6083588897267557e-05,
+      "loss": 3.9243,
+      "step": 1460736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6075202949757037e-05,
+      "loss": 3.925,
+      "step": 1461248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.606681700224652e-05,
+      "loss": 3.9301,
+      "step": 1461760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6058431054736e-05,
+      "loss": 3.9362,
+      "step": 1462272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.605006148602921e-05,
+      "loss": 3.9115,
+      "step": 1462784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.604167553851869e-05,
+      "loss": 3.9192,
+      "step": 1463296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.603328959100817e-05,
+      "loss": 3.9272,
+      "step": 1463808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.602490364349765e-05,
+      "loss": 3.9244,
+      "step": 1464320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.601653407479086e-05,
+      "loss": 3.9101,
+      "step": 1464832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.600814812728034e-05,
+      "loss": 3.9117,
+      "step": 1465344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.599976217976982e-05,
+      "loss": 3.9257,
+      "step": 1465856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.59913762322593e-05,
+      "loss": 3.9306,
+      "step": 1466368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.598299028474878e-05,
+      "loss": 3.928,
+      "step": 1466880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.597462071604199e-05,
+      "loss": 3.9271,
+      "step": 1467392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5966251147335203e-05,
+      "loss": 3.9256,
+      "step": 1467904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5957865199824683e-05,
+      "loss": 3.9246,
+      "step": 1468416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5949479252314163e-05,
+      "loss": 3.9103,
+      "step": 1468928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5941093304803643e-05,
+      "loss": 3.9295,
+      "step": 1469440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5932707357293123e-05,
+      "loss": 3.907,
+      "step": 1469952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5924321409782603e-05,
+      "loss": 3.9193,
+      "step": 1470464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5915935462272083e-05,
+      "loss": 3.9176,
+      "step": 1470976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5907549514761566e-05,
+      "loss": 3.9325,
+      "step": 1471488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5899179946054775e-05,
+      "loss": 3.9238,
+      "step": 1472000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5890793998544255e-05,
+      "loss": 3.9212,
+      "step": 1472512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5882408051033735e-05,
+      "loss": 3.9229,
+      "step": 1473024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5874022103523215e-05,
+      "loss": 3.924,
+      "step": 1473536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5865652534816427e-05,
+      "loss": 3.9306,
+      "step": 1474048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5857266587305907e-05,
+      "loss": 3.9045,
+      "step": 1474560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5848880639795387e-05,
+      "loss": 3.9182,
+      "step": 1475072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5840494692284867e-05,
+      "loss": 3.929,
+      "step": 1475584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.583210874477434e-05,
+      "loss": 3.9269,
+      "step": 1476096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5823739176067556e-05,
+      "loss": 3.9179,
+      "step": 1476608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5815353228557036e-05,
+      "loss": 3.9133,
+      "step": 1477120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5806967281046513e-05,
+      "loss": 3.9183,
+      "step": 1477632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.579859771233973e-05,
+      "loss": 3.9019,
+      "step": 1478144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.579021176482921e-05,
+      "loss": 3.9303,
+      "step": 1478656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.578182581731869e-05,
+      "loss": 3.91,
+      "step": 1479168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5773439869808165e-05,
+      "loss": 3.9205,
+      "step": 1479680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5765053922297645e-05,
+      "loss": 3.9236,
+      "step": 1480192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5756667974787125e-05,
+      "loss": 3.9112,
+      "step": 1480704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5748282027276605e-05,
+      "loss": 3.9139,
+      "step": 1481216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5739912458569814e-05,
+      "loss": 3.9147,
+      "step": 1481728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5731526511059294e-05,
+      "loss": 3.9095,
+      "step": 1482240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5723140563548774e-05,
+      "loss": 3.9182,
+      "step": 1482752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5714754616038254e-05,
+      "loss": 3.9265,
+      "step": 1483264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5706368668527737e-05,
+      "loss": 3.9123,
+      "step": 1483776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5697982721017217e-05,
+      "loss": 3.9139,
+      "step": 1484288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5689596773506697e-05,
+      "loss": 3.9014,
+      "step": 1484800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5681210825996177e-05,
+      "loss": 3.9108,
+      "step": 1485312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5672841257289386e-05,
+      "loss": 3.92,
+      "step": 1485824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5664455309778866e-05,
+      "loss": 3.9347,
+      "step": 1486336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5656069362268346e-05,
+      "loss": 3.9129,
+      "step": 1486848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.564768341475783e-05,
+      "loss": 3.9263,
+      "step": 1487360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.563931384605104e-05,
+      "loss": 3.9299,
+      "step": 1487872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5630944277344248e-05,
+      "loss": 3.9265,
+      "step": 1488384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5622558329833728e-05,
+      "loss": 3.903,
+      "step": 1488896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5614172382323208e-05,
+      "loss": 3.9236,
+      "step": 1489408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.560578643481269e-05,
+      "loss": 3.9191,
+      "step": 1489920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.559740048730217e-05,
+      "loss": 3.9122,
+      "step": 1490432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.558903091859538e-05,
+      "loss": 3.919,
+      "step": 1490944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.558064497108486e-05,
+      "loss": 3.9192,
+      "step": 1491456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.557225902357434e-05,
+      "loss": 3.9318,
+      "step": 1491968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.556387307606382e-05,
+      "loss": 3.9192,
+      "step": 1492480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.55554871285533e-05,
+      "loss": 3.9067,
+      "step": 1492992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5547117559846512e-05,
+      "loss": 3.9155,
+      "step": 1493504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5538731612335992e-05,
+      "loss": 3.9142,
+      "step": 1494016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5530345664825472e-05,
+      "loss": 3.9236,
+      "step": 1494528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5521959717314952e-05,
+      "loss": 3.924,
+      "step": 1495040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5513573769804432e-05,
+      "loss": 3.9215,
+      "step": 1495552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5505187822293912e-05,
+      "loss": 3.9217,
+      "step": 1496064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5496818253587125e-05,
+      "loss": 3.9099,
+      "step": 1496576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5488432306076605e-05,
+      "loss": 3.9107,
+      "step": 1497088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5480046358566085e-05,
+      "loss": 3.9141,
+      "step": 1497600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5471660411055565e-05,
+      "loss": 3.9199,
+      "step": 1498112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5463290842348774e-05,
+      "loss": 3.9198,
+      "step": 1498624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5454904894838254e-05,
+      "loss": 3.928,
+      "step": 1499136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5446518947327737e-05,
+      "loss": 3.9036,
+      "step": 1499648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5438132999817217e-05,
+      "loss": 3.9236,
+      "step": 1500160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5429747052306697e-05,
+      "loss": 3.9293,
+      "step": 1500672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.542136110479617e-05,
+      "loss": 3.9221,
+      "step": 1501184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.541297515728565e-05,
+      "loss": 3.9193,
+      "step": 1501696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.540458920977513e-05,
+      "loss": 3.9072,
+      "step": 1502208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5396219641068346e-05,
+      "loss": 3.9195,
+      "step": 1502720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5387833693557822e-05,
+      "loss": 3.9223,
+      "step": 1503232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5379447746047302e-05,
+      "loss": 3.9039,
+      "step": 1503744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5371078177340518e-05,
+      "loss": 3.912,
+      "step": 1504256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5362692229829998e-05,
+      "loss": 3.9161,
+      "step": 1504768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5354306282319475e-05,
+      "loss": 3.9231,
+      "step": 1505280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5345920334808955e-05,
+      "loss": 3.9048,
+      "step": 1505792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5337534387298435e-05,
+      "loss": 3.927,
+      "step": 1506304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5329148439787915e-05,
+      "loss": 3.9021,
+      "step": 1506816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5320762492277395e-05,
+      "loss": 3.9251,
+      "step": 1507328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5312376544766875e-05,
+      "loss": 3.9141,
+      "step": 1507840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5303990597256355e-05,
+      "loss": 3.9096,
+      "step": 1508352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5295621028549567e-05,
+      "loss": 3.9267,
+      "step": 1508864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5287235081039047e-05,
+      "loss": 3.9292,
+      "step": 1509376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5278865512332256e-05,
+      "loss": 3.9197,
+      "step": 1509888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5270479564821736e-05,
+      "loss": 3.9134,
+      "step": 1510400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5262093617311216e-05,
+      "loss": 3.9175,
+      "step": 1510912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5253707669800696e-05,
+      "loss": 3.921,
+      "step": 1511424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5245321722290176e-05,
+      "loss": 3.9108,
+      "step": 1511936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.523693577477966e-05,
+      "loss": 3.9168,
+      "step": 1512448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.522854982726914e-05,
+      "loss": 3.9146,
+      "step": 1512960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5220180258562348e-05,
+      "loss": 3.9184,
+      "step": 1513472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5211794311051828e-05,
+      "loss": 3.9169,
+      "step": 1513984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5203408363541308e-05,
+      "loss": 3.9165,
+      "step": 1514496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5195022416030788e-05,
+      "loss": 3.9006,
+      "step": 1515008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5186636468520268e-05,
+      "loss": 3.9184,
+      "step": 1515520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5178250521009748e-05,
+      "loss": 3.9116,
+      "step": 1516032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.516986457349923e-05,
+      "loss": 3.9268,
+      "step": 1516544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.516147862598871e-05,
+      "loss": 3.9251,
+      "step": 1517056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.515310905728192e-05,
+      "loss": 3.9223,
+      "step": 1517568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.51447231097714e-05,
+      "loss": 3.9071,
+      "step": 1518080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.513633716226088e-05,
+      "loss": 3.9238,
+      "step": 1518592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.512795121475036e-05,
+      "loss": 3.9215,
+      "step": 1519104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5119581646043573e-05,
+      "loss": 3.9269,
+      "step": 1519616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5111212077336782e-05,
+      "loss": 3.9088,
+      "step": 1520128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5102826129826262e-05,
+      "loss": 3.9283,
+      "step": 1520640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5094440182315742e-05,
+      "loss": 3.9248,
+      "step": 1521152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5086054234805222e-05,
+      "loss": 3.9307,
+      "step": 1521664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5077668287294702e-05,
+      "loss": 3.9101,
+      "step": 1522176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5069282339784185e-05,
+      "loss": 3.923,
+      "step": 1522688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5060912771077394e-05,
+      "loss": 3.919,
+      "step": 1523200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5052526823566874e-05,
+      "loss": 3.9119,
+      "step": 1523712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5044140876056354e-05,
+      "loss": 3.9235,
+      "step": 1524224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.503575492854583e-05,
+      "loss": 3.9283,
+      "step": 1524736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.502736898103531e-05,
+      "loss": 3.9188,
+      "step": 1525248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.501898303352479e-05,
+      "loss": 3.917,
+      "step": 1525760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.501059708601427e-05,
+      "loss": 3.9153,
+      "step": 1526272
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9997053146362305,
+      "eval_runtime": 299.0761,
+      "eval_samples_per_second": 1275.9,
+      "eval_steps_per_second": 39.873,
+      "step": 1526400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.500221113850375e-05,
+      "loss": 3.908,
+      "step": 1526784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4993841569796963e-05,
+      "loss": 3.9084,
+      "step": 1527296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4985472001090175e-05,
+      "loss": 3.9241,
+      "step": 1527808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4977086053579655e-05,
+      "loss": 3.9169,
+      "step": 1528320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4968700106069135e-05,
+      "loss": 3.9287,
+      "step": 1528832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4960314158558615e-05,
+      "loss": 3.9143,
+      "step": 1529344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4951928211048095e-05,
+      "loss": 3.9159,
+      "step": 1529856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4943542263537572e-05,
+      "loss": 3.9023,
+      "step": 1530368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4935156316027052e-05,
+      "loss": 3.9204,
+      "step": 1530880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4926770368516532e-05,
+      "loss": 3.9183,
+      "step": 1531392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4918400799809748e-05,
+      "loss": 3.9231,
+      "step": 1531904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4910014852299224e-05,
+      "loss": 3.9249,
+      "step": 1532416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4901645283592437e-05,
+      "loss": 3.9104,
+      "step": 1532928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4893259336081917e-05,
+      "loss": 3.912,
+      "step": 1533440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4884873388571393e-05,
+      "loss": 3.9097,
+      "step": 1533952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4876487441060877e-05,
+      "loss": 3.9045,
+      "step": 1534464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4868101493550356e-05,
+      "loss": 3.9133,
+      "step": 1534976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4859715546039836e-05,
+      "loss": 3.9109,
+      "step": 1535488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4851329598529316e-05,
+      "loss": 3.9106,
+      "step": 1536000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4842943651018796e-05,
+      "loss": 3.9343,
+      "step": 1536512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4834574082312005e-05,
+      "loss": 3.9169,
+      "step": 1537024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4826188134801485e-05,
+      "loss": 3.9143,
+      "step": 1537536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.481780218729097e-05,
+      "loss": 3.9245,
+      "step": 1538048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.480941623978045e-05,
+      "loss": 3.9269,
+      "step": 1538560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4801046671073658e-05,
+      "loss": 3.9062,
+      "step": 1539072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4792660723563138e-05,
+      "loss": 3.9164,
+      "step": 1539584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4784274776052618e-05,
+      "loss": 3.9168,
+      "step": 1540096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4775888828542098e-05,
+      "loss": 3.9158,
+      "step": 1540608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.476751925983531e-05,
+      "loss": 3.9016,
+      "step": 1541120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.475913331232479e-05,
+      "loss": 3.9066,
+      "step": 1541632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.475074736481427e-05,
+      "loss": 3.9199,
+      "step": 1542144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.474236141730375e-05,
+      "loss": 3.923,
+      "step": 1542656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4733975469793227e-05,
+      "loss": 3.9187,
+      "step": 1543168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.472560590108644e-05,
+      "loss": 3.9233,
+      "step": 1543680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.471723633237965e-05,
+      "loss": 3.9165,
+      "step": 1544192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.470885038486913e-05,
+      "loss": 3.9209,
+      "step": 1544704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.470046443735861e-05,
+      "loss": 3.9073,
+      "step": 1545216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.469207848984809e-05,
+      "loss": 3.9168,
+      "step": 1545728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.468369254233757e-05,
+      "loss": 3.8998,
+      "step": 1546240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4675306594827048e-05,
+      "loss": 3.9131,
+      "step": 1546752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.466692064731653e-05,
+      "loss": 3.9101,
+      "step": 1547264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.465853469980601e-05,
+      "loss": 3.9282,
+      "step": 1547776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4650165131099224e-05,
+      "loss": 3.9156,
+      "step": 1548288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.46417791835887e-05,
+      "loss": 3.915,
+      "step": 1548800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.463339323607818e-05,
+      "loss": 3.9146,
+      "step": 1549312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.462500728856766e-05,
+      "loss": 3.9177,
+      "step": 1549824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4616637719860876e-05,
+      "loss": 3.9237,
+      "step": 1550336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4608251772350353e-05,
+      "loss": 3.9,
+      "step": 1550848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4599882203643565e-05,
+      "loss": 3.9068,
+      "step": 1551360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4591496256133045e-05,
+      "loss": 3.9248,
+      "step": 1551872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.458311030862252e-05,
+      "loss": 3.9201,
+      "step": 1552384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4574724361112e-05,
+      "loss": 3.9082,
+      "step": 1552896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4566354792405217e-05,
+      "loss": 3.9051,
+      "step": 1553408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4557968844894697e-05,
+      "loss": 3.9138,
+      "step": 1553920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4549582897384174e-05,
+      "loss": 3.8982,
+      "step": 1554432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4541213328677386e-05,
+      "loss": 3.9189,
+      "step": 1554944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4532827381166866e-05,
+      "loss": 3.9036,
+      "step": 1555456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4524441433656346e-05,
+      "loss": 3.9116,
+      "step": 1555968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4516055486145826e-05,
+      "loss": 3.9132,
+      "step": 1556480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4507669538635306e-05,
+      "loss": 3.91,
+      "step": 1556992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4499283591124786e-05,
+      "loss": 3.9013,
+      "step": 1557504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4490897643614266e-05,
+      "loss": 3.9092,
+      "step": 1558016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4482511696103746e-05,
+      "loss": 3.9004,
+      "step": 1558528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4474125748593226e-05,
+      "loss": 3.9101,
+      "step": 1559040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4465739801082703e-05,
+      "loss": 3.918,
+      "step": 1559552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.445737023237592e-05,
+      "loss": 3.9075,
+      "step": 1560064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.44489842848654e-05,
+      "loss": 3.9056,
+      "step": 1560576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.444059833735488e-05,
+      "loss": 3.8942,
+      "step": 1561088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4432212389844355e-05,
+      "loss": 3.9008,
+      "step": 1561600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4423826442333835e-05,
+      "loss": 3.9158,
+      "step": 1562112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4415440494823315e-05,
+      "loss": 3.9223,
+      "step": 1562624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4407054547312795e-05,
+      "loss": 3.9093,
+      "step": 1563136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4398668599802278e-05,
+      "loss": 3.9178,
+      "step": 1563648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4390282652291758e-05,
+      "loss": 3.9256,
+      "step": 1564160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4381913083584967e-05,
+      "loss": 3.9177,
+      "step": 1564672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4373527136074447e-05,
+      "loss": 3.8977,
+      "step": 1565184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4365141188563927e-05,
+      "loss": 3.9141,
+      "step": 1565696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.435677161985714e-05,
+      "loss": 3.9124,
+      "step": 1566208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.434838567234662e-05,
+      "loss": 3.9041,
+      "step": 1566720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.43399997248361e-05,
+      "loss": 3.909,
+      "step": 1567232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.433161377732558e-05,
+      "loss": 3.9155,
+      "step": 1567744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.432322782981506e-05,
+      "loss": 3.9265,
+      "step": 1568256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4314841882304536e-05,
+      "loss": 3.9118,
+      "step": 1568768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4306455934794016e-05,
+      "loss": 3.8965,
+      "step": 1569280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.42980699872835e-05,
+      "loss": 3.9102,
+      "step": 1569792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.428970041857671e-05,
+      "loss": 3.9106,
+      "step": 1570304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.428133084986992e-05,
+      "loss": 3.9112,
+      "step": 1570816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.42729449023594e-05,
+      "loss": 3.9162,
+      "step": 1571328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.426455895484888e-05,
+      "loss": 3.914,
+      "step": 1571840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.425617300733836e-05,
+      "loss": 3.9151,
+      "step": 1572352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.424778705982784e-05,
+      "loss": 3.9054,
+      "step": 1572864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.423940111231732e-05,
+      "loss": 3.9013,
+      "step": 1573376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.42310151648068e-05,
+      "loss": 3.9056,
+      "step": 1573888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.422262921729628e-05,
+      "loss": 3.917,
+      "step": 1574400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.421425964858949e-05,
+      "loss": 3.9143,
+      "step": 1574912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.420587370107897e-05,
+      "loss": 3.9189,
+      "step": 1575424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.419748775356845e-05,
+      "loss": 3.8981,
+      "step": 1575936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4189101806057933e-05,
+      "loss": 3.9138,
+      "step": 1576448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4180715858547413e-05,
+      "loss": 3.9204,
+      "step": 1576960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4172346289840622e-05,
+      "loss": 3.9178,
+      "step": 1577472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4163960342330102e-05,
+      "loss": 3.9111,
+      "step": 1577984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4155574394819582e-05,
+      "loss": 3.904,
+      "step": 1578496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4147188447309062e-05,
+      "loss": 3.9089,
+      "step": 1579008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4138818878602274e-05,
+      "loss": 3.9179,
+      "step": 1579520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4130432931091754e-05,
+      "loss": 3.8928,
+      "step": 1580032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4122063362384963e-05,
+      "loss": 3.9092,
+      "step": 1580544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4113677414874443e-05,
+      "loss": 3.9054,
+      "step": 1581056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4105291467363923e-05,
+      "loss": 3.9152,
+      "step": 1581568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4096905519853403e-05,
+      "loss": 3.8992,
+      "step": 1582080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4088519572342887e-05,
+      "loss": 3.9189,
+      "step": 1582592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4080133624832363e-05,
+      "loss": 3.9054,
+      "step": 1583104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4071747677321843e-05,
+      "loss": 3.9155,
+      "step": 1583616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4063361729811323e-05,
+      "loss": 3.9061,
+      "step": 1584128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4054992161104536e-05,
+      "loss": 3.8987,
+      "step": 1584640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4046606213594016e-05,
+      "loss": 3.9182,
+      "step": 1585152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4038220266083496e-05,
+      "loss": 3.9228,
+      "step": 1585664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4029850697376708e-05,
+      "loss": 3.9154,
+      "step": 1586176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4021464749866188e-05,
+      "loss": 3.9042,
+      "step": 1586688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4013078802355665e-05,
+      "loss": 3.9109,
+      "step": 1587200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4004709233648877e-05,
+      "loss": 3.9145,
+      "step": 1587712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3996323286138357e-05,
+      "loss": 3.9014,
+      "step": 1588224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3987937338627837e-05,
+      "loss": 3.9115,
+      "step": 1588736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3979551391117317e-05,
+      "loss": 3.9051,
+      "step": 1589248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3971165443606797e-05,
+      "loss": 3.9097,
+      "step": 1589760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3962779496096277e-05,
+      "loss": 3.9099,
+      "step": 1590272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3954393548585757e-05,
+      "loss": 3.9123,
+      "step": 1590784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3946007601075237e-05,
+      "loss": 3.8961,
+      "step": 1591296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3937621653564717e-05,
+      "loss": 3.9104,
+      "step": 1591808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.392925208485793e-05,
+      "loss": 3.9049,
+      "step": 1592320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.392086613734741e-05,
+      "loss": 3.9208,
+      "step": 1592832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.391248018983689e-05,
+      "loss": 3.9119,
+      "step": 1593344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.390409424232637e-05,
+      "loss": 3.9182,
+      "step": 1593856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3895708294815846e-05,
+      "loss": 3.8995,
+      "step": 1594368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3887322347305326e-05,
+      "loss": 3.9184,
+      "step": 1594880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.387895277859854e-05,
+      "loss": 3.9168,
+      "step": 1595392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3870566831088018e-05,
+      "loss": 3.9163,
+      "step": 1595904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3862180883577498e-05,
+      "loss": 3.9021,
+      "step": 1596416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3853794936066978e-05,
+      "loss": 3.9175,
+      "step": 1596928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3845408988556458e-05,
+      "loss": 3.9184,
+      "step": 1597440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3837023041045938e-05,
+      "loss": 3.9272,
+      "step": 1597952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3828637093535418e-05,
+      "loss": 3.9012,
+      "step": 1598464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.382026752482863e-05,
+      "loss": 3.9161,
+      "step": 1598976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.381188157731811e-05,
+      "loss": 3.9131,
+      "step": 1599488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.380349562980759e-05,
+      "loss": 3.9035,
+      "step": 1600000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.379510968229707e-05,
+      "loss": 3.9183,
+      "step": 1600512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3786723734786547e-05,
+      "loss": 3.9198,
+      "step": 1601024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3778337787276027e-05,
+      "loss": 3.915,
+      "step": 1601536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.376995183976551e-05,
+      "loss": 3.9056,
+      "step": 1602048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3761582271058722e-05,
+      "loss": 3.9091,
+      "step": 1602560
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.996756076812744,
+      "eval_runtime": 305.7296,
+      "eval_samples_per_second": 1248.132,
+      "eval_steps_per_second": 39.005,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.37531963235482e-05,
+      "loss": 3.8985,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.374481037603768e-05,
+      "loss": 3.8997,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.373642442852716e-05,
+      "loss": 3.9148,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.372803848101664e-05,
+      "loss": 3.9106,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.371965253350612e-05,
+      "loss": 3.9182,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3711266585995602e-05,
+      "loss": 3.9097,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3702880638485082e-05,
+      "loss": 3.9063,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.369449469097456e-05,
+      "loss": 3.898,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.368610874346404e-05,
+      "loss": 3.9099,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.367772279595352e-05,
+      "loss": 3.9116,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3669336848443e-05,
+      "loss": 3.9152,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.366095090093248e-05,
+      "loss": 3.9177,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.365256495342196e-05,
+      "loss": 3.9033,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.364417900591144e-05,
+      "loss": 3.9066,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.363579305840092e-05,
+      "loss": 3.9059,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.36274071108904e-05,
+      "loss": 3.8951,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.361902116337988e-05,
+      "loss": 3.9086,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3610651594673087e-05,
+      "loss": 3.9036,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.360226564716257e-05,
+      "loss": 3.9013,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.359387969965205e-05,
+      "loss": 3.9222,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.358549375214153e-05,
+      "loss": 3.9147,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.357710780463101e-05,
+      "loss": 3.9111,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356872185712049e-05,
+      "loss": 3.9156,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356033590960997e-05,
+      "loss": 3.9203,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3551949962099447e-05,
+      "loss": 3.896,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3543580393392663e-05,
+      "loss": 3.9088,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3535194445882143e-05,
+      "loss": 3.9152,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.352680849837162e-05,
+      "loss": 3.9032,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.35184225508611e-05,
+      "loss": 3.8967,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3510052982154312e-05,
+      "loss": 3.899,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3501683413447524e-05,
+      "loss": 3.9138,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3493297465937004e-05,
+      "loss": 3.9129,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3484911518426484e-05,
+      "loss": 3.9132,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3476525570915964e-05,
+      "loss": 3.9166,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3468156002209173e-05,
+      "loss": 3.9102,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3459770054698653e-05,
+      "loss": 3.9096,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3451384107188133e-05,
+      "loss": 3.9045,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3442998159677613e-05,
+      "loss": 3.9097,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3434612212167093e-05,
+      "loss": 3.8935,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3426226264656573e-05,
+      "loss": 3.9021,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3417856695949786e-05,
+      "loss": 3.9033,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3409470748439266e-05,
+      "loss": 3.9193,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3401084800928742e-05,
+      "loss": 3.9105,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3392698853418226e-05,
+      "loss": 3.9093,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3384312905907705e-05,
+      "loss": 3.9125,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3375926958397185e-05,
+      "loss": 3.9053,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3367541010886665e-05,
+      "loss": 3.9164,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3359155063376145e-05,
+      "loss": 3.8987,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3350785494669354e-05,
+      "loss": 3.8975,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3342415925962567e-05,
+      "loss": 3.9157,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3334029978452047e-05,
+      "loss": 3.9113,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3325644030941527e-05,
+      "loss": 3.9007,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3317258083431007e-05,
+      "loss": 3.902,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3308872135920487e-05,
+      "loss": 3.908,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3300486188409967e-05,
+      "loss": 3.8902,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.329211661970318e-05,
+      "loss": 3.9128,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.328373067219266e-05,
+      "loss": 3.8988,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.327534472468214e-05,
+      "loss": 3.9022,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.326695877717162e-05,
+      "loss": 3.9101,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3258572829661096e-05,
+      "loss": 3.9017,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3250186882150576e-05,
+      "loss": 3.8993,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3241800934640056e-05,
+      "loss": 3.9024,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3233414987129536e-05,
+      "loss": 3.894,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3225045418422748e-05,
+      "loss": 3.9008,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3216659470912228e-05,
+      "loss": 3.9117,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3208273523401708e-05,
+      "loss": 3.9034,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3199887575891188e-05,
+      "loss": 3.898,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3191518007184397e-05,
+      "loss": 3.8917,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.318313205967388e-05,
+      "loss": 3.8923,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.317474611216336e-05,
+      "loss": 3.9056,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.316636016465284e-05,
+      "loss": 3.9135,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.315799059594605e-05,
+      "loss": 3.9054,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3149621027239262e-05,
+      "loss": 3.9103,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3141235079728742e-05,
+      "loss": 3.9146,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3132849132218222e-05,
+      "loss": 3.9173,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.31244631847077e-05,
+      "loss": 3.89,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.311607723719718e-05,
+      "loss": 3.9065,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.310769128968666e-05,
+      "loss": 3.9097,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309930534217614e-05,
+      "loss": 3.8936,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309091939466562e-05,
+      "loss": 3.9073,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.30825334471551e-05,
+      "loss": 3.9032,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3074180257252043e-05,
+      "loss": 3.9176,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3065794309741523e-05,
+      "loss": 3.9068,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3057408362231003e-05,
+      "loss": 3.8882,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3049022414720483e-05,
+      "loss": 3.9028,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3040636467209963e-05,
+      "loss": 3.9026,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3032250519699443e-05,
+      "loss": 3.9089,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3023880950992655e-05,
+      "loss": 3.9057,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3015495003482135e-05,
+      "loss": 3.9121,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3007109055971615e-05,
+      "loss": 3.9072,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2998723108461095e-05,
+      "loss": 3.8997,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2990337160950575e-05,
+      "loss": 3.8927,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2981951213440052e-05,
+      "loss": 3.9006,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2973565265929535e-05,
+      "loss": 3.9113,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2965179318419015e-05,
+      "loss": 3.9073,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2956809749712224e-05,
+      "loss": 3.9103,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2948423802201704e-05,
+      "loss": 3.8924,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2940037854691184e-05,
+      "loss": 3.9064,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2931651907180664e-05,
+      "loss": 3.9133,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2923265959670144e-05,
+      "loss": 3.9117,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2914896390963356e-05,
+      "loss": 3.902,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.290652682225657e-05,
+      "loss": 3.8987,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.289814087474605e-05,
+      "loss": 3.8998,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2889754927235525e-05,
+      "loss": 3.9126,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2881368979725005e-05,
+      "loss": 3.8835,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.287298303221449e-05,
+      "loss": 3.9041,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.286459708470397e-05,
+      "loss": 3.8945,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.285621113719345e-05,
+      "loss": 3.91,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2847841568486658e-05,
+      "loss": 3.8911,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2839455620976138e-05,
+      "loss": 3.915,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2831069673465618e-05,
+      "loss": 3.8984,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2822683725955098e-05,
+      "loss": 3.9091,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.281429777844458e-05,
+      "loss": 3.8996,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2805911830934058e-05,
+      "loss": 3.8964,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2797525883423537e-05,
+      "loss": 3.9079,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2789139935913017e-05,
+      "loss": 3.9173,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.278078674600996e-05,
+      "loss": 3.9064,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2772400798499442e-05,
+      "loss": 3.8954,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2764014850988922e-05,
+      "loss": 3.9041,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2755628903478402e-05,
+      "loss": 3.9097,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274724295596788e-05,
+      "loss": 3.8957,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.273885700845736e-05,
+      "loss": 3.909,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.273047106094684e-05,
+      "loss": 3.893,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.272210149224005e-05,
+      "loss": 3.9033,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.271371554472953e-05,
+      "loss": 3.9044,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.270532959721901e-05,
+      "loss": 3.9067,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.269694364970849e-05,
+      "loss": 3.8859,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268855770219797e-05,
+      "loss": 3.9063,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268018813349118e-05,
+      "loss": 3.8963,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.267180218598066e-05,
+      "loss": 3.9131,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2663416238470143e-05,
+      "loss": 3.9085,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2655030290959623e-05,
+      "loss": 3.9109,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2646644343449103e-05,
+      "loss": 3.8933,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2638258395938583e-05,
+      "loss": 3.9127,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262987244842806e-05,
+      "loss": 3.9095,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262148650091754e-05,
+      "loss": 3.9073,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2613116932210752e-05,
+      "loss": 3.8991,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2604730984700236e-05,
+      "loss": 3.9096,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2596345037189712e-05,
+      "loss": 3.9133,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2587959089679192e-05,
+      "loss": 3.9176,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2579589520972405e-05,
+      "loss": 3.8949,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257120357346188e-05,
+      "loss": 3.9094,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2562817625951365e-05,
+      "loss": 3.9057,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2554431678440845e-05,
+      "loss": 3.897,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2546045730930325e-05,
+      "loss": 3.9114,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2537676162223534e-05,
+      "loss": 3.9068,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2529306593516746e-05,
+      "loss": 3.9152,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2520920646006226e-05,
+      "loss": 3.8977,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2512534698495706e-05,
+      "loss": 3.9025,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.994250535964966,
+      "eval_runtime": 291.9721,
+      "eval_samples_per_second": 1306.943,
+      "eval_steps_per_second": 40.843,
+      "step": 1679040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2504148750985186e-05,
+      "loss": 3.899,
+      "step": 1679360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2495762803474666e-05,
+      "loss": 3.8895,
+      "step": 1679872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2487376855964146e-05,
+      "loss": 3.9089,
+      "step": 1680384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2478990908453626e-05,
+      "loss": 3.9045,
+      "step": 1680896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2470604960943106e-05,
+      "loss": 3.9133,
+      "step": 1681408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2462235392236315e-05,
+      "loss": 3.9036,
+      "step": 1681920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2453849444725798e-05,
+      "loss": 3.8946,
+      "step": 1682432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2445463497215278e-05,
+      "loss": 3.897,
+      "step": 1682944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2437077549704758e-05,
+      "loss": 3.9023,
+      "step": 1683456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2428691602194238e-05,
+      "loss": 3.9037,
+      "step": 1683968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2420322033487447e-05,
+      "loss": 3.911,
+      "step": 1684480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2411936085976927e-05,
+      "loss": 3.9132,
+      "step": 1684992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.240356651727014e-05,
+      "loss": 3.8981,
+      "step": 1685504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.239518056975962e-05,
+      "loss": 3.8972,
+      "step": 1686016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.23867946222491e-05,
+      "loss": 3.9032,
+      "step": 1686528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.237840867473858e-05,
+      "loss": 3.886,
+      "step": 1687040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.237003910603179e-05,
+      "loss": 3.901,
+      "step": 1687552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.236165315852127e-05,
+      "loss": 3.8959,
+      "step": 1688064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2353267211010752e-05,
+      "loss": 3.8968,
+      "step": 1688576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2344881263500232e-05,
+      "loss": 3.9134,
+      "step": 1689088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2336495315989712e-05,
+      "loss": 3.908,
+      "step": 1689600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.232812574728292e-05,
+      "loss": 3.9063,
+      "step": 1690112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.23197397997724e-05,
+      "loss": 3.9044,
+      "step": 1690624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.231135385226188e-05,
+      "loss": 3.9135,
+      "step": 1691136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.230296790475136e-05,
+      "loss": 3.8946,
+      "step": 1691648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.229458195724084e-05,
+      "loss": 3.9042,
+      "step": 1692160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.228619600973032e-05,
+      "loss": 3.9002,
+      "step": 1692672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.22778100622198e-05,
+      "loss": 3.8972,
+      "step": 1693184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.226942411470928e-05,
+      "loss": 3.8948,
+      "step": 1693696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.226105454600249e-05,
+      "loss": 3.8883,
+      "step": 1694208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2252668598491973e-05,
+      "loss": 3.9088,
+      "step": 1694720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2244282650981453e-05,
+      "loss": 3.9065,
+      "step": 1695232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2235913082274662e-05,
+      "loss": 3.9064,
+      "step": 1695744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2227527134764142e-05,
+      "loss": 3.9116,
+      "step": 1696256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2219157566057355e-05,
+      "loss": 3.9035,
+      "step": 1696768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2210771618546834e-05,
+      "loss": 3.9052,
+      "step": 1697280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2202385671036314e-05,
+      "loss": 3.9004,
+      "step": 1697792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2193999723525794e-05,
+      "loss": 3.8981,
+      "step": 1698304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2185613776015274e-05,
+      "loss": 3.8879,
+      "step": 1698816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2177227828504754e-05,
+      "loss": 3.8947,
+      "step": 1699328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2168841880994234e-05,
+      "loss": 3.8995,
+      "step": 1699840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2160455933483714e-05,
+      "loss": 3.9139,
+      "step": 1700352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.215206998597319e-05,
+      "loss": 3.9019,
+      "step": 1700864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2143684038462674e-05,
+      "loss": 3.9001,
+      "step": 1701376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2135298090952154e-05,
+      "loss": 3.909,
+      "step": 1701888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2126912143441634e-05,
+      "loss": 3.8995,
+      "step": 1702400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2118542574734843e-05,
+      "loss": 3.9109,
+      "step": 1702912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2110173006028056e-05,
+      "loss": 3.893,
+      "step": 1703424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2101787058517536e-05,
+      "loss": 3.8853,
+      "step": 1703936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2093401111007016e-05,
+      "loss": 3.9103,
+      "step": 1704448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2085015163496495e-05,
+      "loss": 3.9062,
+      "step": 1704960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2076629215985975e-05,
+      "loss": 3.899,
+      "step": 1705472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2068259647279188e-05,
+      "loss": 3.8922,
+      "step": 1705984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2059873699768665e-05,
+      "loss": 3.9025,
+      "step": 1706496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2051487752258144e-05,
+      "loss": 3.8825,
+      "step": 1707008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2043101804747628e-05,
+      "loss": 3.905,
+      "step": 1707520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2034715857237108e-05,
+      "loss": 3.8947,
+      "step": 1708032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2026329909726588e-05,
+      "loss": 3.8927,
+      "step": 1708544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2017943962216068e-05,
+      "loss": 3.905,
+      "step": 1709056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2009558014705548e-05,
+      "loss": 3.8921,
+      "step": 1709568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2001188445998757e-05,
+      "loss": 3.8937,
+      "step": 1710080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1992802498488237e-05,
+      "loss": 3.8994,
+      "step": 1710592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.198441655097772e-05,
+      "loss": 3.8857,
+      "step": 1711104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1976030603467197e-05,
+      "loss": 3.8965,
+      "step": 1711616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1967644655956677e-05,
+      "loss": 3.9038,
+      "step": 1712128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1959258708446157e-05,
+      "loss": 3.8979,
+      "step": 1712640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1950872760935636e-05,
+      "loss": 3.8904,
+      "step": 1713152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1942486813425116e-05,
+      "loss": 3.8861,
+      "step": 1713664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.193411724471833e-05,
+      "loss": 3.8823,
+      "step": 1714176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.192573129720781e-05,
+      "loss": 3.8963,
+      "step": 1714688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.191734534969729e-05,
+      "loss": 3.9079,
+      "step": 1715200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.190895940218677e-05,
+      "loss": 3.8973,
+      "step": 1715712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.190057345467625e-05,
+      "loss": 3.9051,
+      "step": 1716224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.189218750716573e-05,
+      "loss": 3.9062,
+      "step": 1716736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1883801559655205e-05,
+      "loss": 3.9107,
+      "step": 1717248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1875415612144685e-05,
+      "loss": 3.8847,
+      "step": 1717760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.18670460434379e-05,
+      "loss": 3.902,
+      "step": 1718272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1858660095927378e-05,
+      "loss": 3.9009,
+      "step": 1718784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1850274148416858e-05,
+      "loss": 3.8832,
+      "step": 1719296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1841888200906338e-05,
+      "loss": 3.9052,
+      "step": 1719808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.183351863219955e-05,
+      "loss": 3.8982,
+      "step": 1720320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.182513268468903e-05,
+      "loss": 3.9077,
+      "step": 1720832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.181674673717851e-05,
+      "loss": 3.9018,
+      "step": 1721344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1808377168471722e-05,
+      "loss": 3.8822,
+      "step": 1721856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1799991220961202e-05,
+      "loss": 3.8948,
+      "step": 1722368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.179160527345068e-05,
+      "loss": 3.8963,
+      "step": 1722880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.178321932594016e-05,
+      "loss": 3.9003,
+      "step": 1723392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.177483337842964e-05,
+      "loss": 3.8987,
+      "step": 1723904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.176646380972285e-05,
+      "loss": 3.9063,
+      "step": 1724416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.175807786221233e-05,
+      "loss": 3.8991,
+      "step": 1724928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.174969191470181e-05,
+      "loss": 3.8959,
+      "step": 1725440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.174130596719129e-05,
+      "loss": 3.8872,
+      "step": 1725952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.173292001968077e-05,
+      "loss": 3.8896,
+      "step": 1726464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.172453407217025e-05,
+      "loss": 3.9086,
+      "step": 1726976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.171614812465973e-05,
+      "loss": 3.9017,
+      "step": 1727488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1707778555952944e-05,
+      "loss": 3.9013,
+      "step": 1728000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1699392608442424e-05,
+      "loss": 3.8873,
+      "step": 1728512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1691006660931903e-05,
+      "loss": 3.898,
+      "step": 1729024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1682620713421383e-05,
+      "loss": 3.906,
+      "step": 1729536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.167423476591086e-05,
+      "loss": 3.9037,
+      "step": 1730048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.166584881840034e-05,
+      "loss": 3.8987,
+      "step": 1730560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1657479249693556e-05,
+      "loss": 3.8927,
+      "step": 1731072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1649093302183032e-05,
+      "loss": 3.8953,
+      "step": 1731584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1640707354672512e-05,
+      "loss": 3.904,
+      "step": 1732096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1632321407161992e-05,
+      "loss": 3.8783,
+      "step": 1732608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1623951838455205e-05,
+      "loss": 3.8916,
+      "step": 1733120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1615565890944685e-05,
+      "loss": 3.8885,
+      "step": 1733632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1607179943434165e-05,
+      "loss": 3.9031,
+      "step": 1734144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1598793995923645e-05,
+      "loss": 3.8891,
+      "step": 1734656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1590408048413125e-05,
+      "loss": 3.9061,
+      "step": 1735168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1582022100902605e-05,
+      "loss": 3.8935,
+      "step": 1735680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1573636153392085e-05,
+      "loss": 3.9,
+      "step": 1736192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1565250205881564e-05,
+      "loss": 3.8983,
+      "step": 1736704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1556864258371044e-05,
+      "loss": 3.8871,
+      "step": 1737216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1548478310860524e-05,
+      "loss": 3.8999,
+      "step": 1737728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1540092363350004e-05,
+      "loss": 3.9098,
+      "step": 1738240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1531722794643213e-05,
+      "loss": 3.8997,
+      "step": 1738752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1523336847132693e-05,
+      "loss": 3.8976,
+      "step": 1739264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1514950899622173e-05,
+      "loss": 3.8884,
+      "step": 1739776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1506564952111653e-05,
+      "loss": 3.9031,
+      "step": 1740288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1498179004601137e-05,
+      "loss": 3.8946,
+      "step": 1740800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1489793057090617e-05,
+      "loss": 3.8966,
+      "step": 1741312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1481407109580097e-05,
+      "loss": 3.8864,
+      "step": 1741824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1473021162069573e-05,
+      "loss": 3.8959,
+      "step": 1742336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1464635214559053e-05,
+      "loss": 3.8993,
+      "step": 1742848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1456265645852266e-05,
+      "loss": 3.8983,
+      "step": 1743360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1447896077145478e-05,
+      "loss": 3.8769,
+      "step": 1743872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1439510129634958e-05,
+      "loss": 3.9001,
+      "step": 1744384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1431124182124438e-05,
+      "loss": 3.8881,
+      "step": 1744896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1422738234613918e-05,
+      "loss": 3.9097,
+      "step": 1745408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1414352287103394e-05,
+      "loss": 3.8991,
+      "step": 1745920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1405966339592874e-05,
+      "loss": 3.9048,
+      "step": 1746432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1397596770886087e-05,
+      "loss": 3.8865,
+      "step": 1746944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.138921082337557e-05,
+      "loss": 3.9078,
+      "step": 1747456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1380824875865047e-05,
+      "loss": 3.9016,
+      "step": 1747968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1372438928354527e-05,
+      "loss": 3.904,
+      "step": 1748480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1364052980844007e-05,
+      "loss": 3.8901,
+      "step": 1748992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1355667033333487e-05,
+      "loss": 3.9041,
+      "step": 1749504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1347281085822967e-05,
+      "loss": 3.9061,
+      "step": 1750016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1338895138312447e-05,
+      "loss": 3.911,
+      "step": 1750528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.133052556960566e-05,
+      "loss": 3.8923,
+      "step": 1751040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.132213962209514e-05,
+      "loss": 3.9016,
+      "step": 1751552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.131375367458462e-05,
+      "loss": 3.9012,
+      "step": 1752064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.13053677270741e-05,
+      "loss": 3.8934,
+      "step": 1752576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1296981779563576e-05,
+      "loss": 3.8987,
+      "step": 1753088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.128861221085679e-05,
+      "loss": 3.9013,
+      "step": 1753600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.128022626334627e-05,
+      "loss": 3.9056,
+      "step": 1754112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.127184031583575e-05,
+      "loss": 3.8946,
+      "step": 1754624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1263454368325228e-05,
+      "loss": 3.8967,
+      "step": 1755136
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9918885231018066,
+      "eval_runtime": 302.7182,
+      "eval_samples_per_second": 1260.548,
+      "eval_steps_per_second": 39.393,
+      "step": 1755360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1255068420814708e-05,
+      "loss": 3.8917,
+      "step": 1755648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1246682473304188e-05,
+      "loss": 3.8843,
+      "step": 1756160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1238296525793668e-05,
+      "loss": 3.9006,
+      "step": 1756672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1229910578283148e-05,
+      "loss": 3.8999,
+      "step": 1757184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.122152463077263e-05,
+      "loss": 3.903,
+      "step": 1757696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1213138683262108e-05,
+      "loss": 3.8987,
+      "step": 1758208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1204752735751588e-05,
+      "loss": 3.8908,
+      "step": 1758720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1196366788241068e-05,
+      "loss": 3.8884,
+      "step": 1759232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1187980840730547e-05,
+      "loss": 3.8962,
+      "step": 1759744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1179594893220027e-05,
+      "loss": 3.8972,
+      "step": 1760256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1171208945709507e-05,
+      "loss": 3.9022,
+      "step": 1760768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1162822998198987e-05,
+      "loss": 3.905,
+      "step": 1761280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1154437050688467e-05,
+      "loss": 3.8932,
+      "step": 1761792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.114606748198168e-05,
+      "loss": 3.8932,
+      "step": 1762304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.113768153447116e-05,
+      "loss": 3.8928,
+      "step": 1762816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.112929558696064e-05,
+      "loss": 3.8779,
+      "step": 1763328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1120909639450116e-05,
+      "loss": 3.895,
+      "step": 1763840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.11125236919396e-05,
+      "loss": 3.892,
+      "step": 1764352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.110413774442908e-05,
+      "loss": 3.893,
+      "step": 1764864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.109575179691856e-05,
+      "loss": 3.9037,
+      "step": 1765376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.108736584940804e-05,
+      "loss": 3.9056,
+      "step": 1765888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.107897990189752e-05,
+      "loss": 3.8966,
+      "step": 1766400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1070593954386996e-05,
+      "loss": 3.8997,
+      "step": 1766912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1062208006876476e-05,
+      "loss": 3.9047,
+      "step": 1767424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1053838438169692e-05,
+      "loss": 3.8869,
+      "step": 1767936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1045452490659172e-05,
+      "loss": 3.897,
+      "step": 1768448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.103706654314865e-05,
+      "loss": 3.8992,
+      "step": 1768960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.102868059563813e-05,
+      "loss": 3.8913,
+      "step": 1769472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1020294648127608e-05,
+      "loss": 3.8866,
+      "step": 1769984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1011908700617088e-05,
+      "loss": 3.8857,
+      "step": 1770496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1003522753106568e-05,
+      "loss": 3.8963,
+      "step": 1771008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0995136805596048e-05,
+      "loss": 3.9034,
+      "step": 1771520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.098675085808553e-05,
+      "loss": 3.8973,
+      "step": 1772032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.097838128937874e-05,
+      "loss": 3.9042,
+      "step": 1772544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.097001172067195e-05,
+      "loss": 3.8944,
+      "step": 1773056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.096162577316143e-05,
+      "loss": 3.8993,
+      "step": 1773568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.095323982565091e-05,
+      "loss": 3.8989,
+      "step": 1774080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0944853878140393e-05,
+      "loss": 3.8853,
+      "step": 1774592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0936467930629873e-05,
+      "loss": 3.8854,
+      "step": 1775104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0928098361923082e-05,
+      "loss": 3.8867,
+      "step": 1775616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0919712414412562e-05,
+      "loss": 3.8912,
+      "step": 1776128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0911326466902042e-05,
+      "loss": 3.908,
+      "step": 1776640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0902940519391522e-05,
+      "loss": 3.8957,
+      "step": 1777152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0894554571881002e-05,
+      "loss": 3.8907,
+      "step": 1777664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0886168624370482e-05,
+      "loss": 3.906,
+      "step": 1778176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0877782676859962e-05,
+      "loss": 3.8939,
+      "step": 1778688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0869413108153174e-05,
+      "loss": 3.9009,
+      "step": 1779200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.086102716064265e-05,
+      "loss": 3.8894,
+      "step": 1779712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.085264121313213e-05,
+      "loss": 3.8769,
+      "step": 1780224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.084425526562161e-05,
+      "loss": 3.9082,
+      "step": 1780736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0835869318111094e-05,
+      "loss": 3.8965,
+      "step": 1781248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0827483370600574e-05,
+      "loss": 3.8939,
+      "step": 1781760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0819113801893783e-05,
+      "loss": 3.8893,
+      "step": 1782272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0810727854383263e-05,
+      "loss": 3.8865,
+      "step": 1782784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0802341906872743e-05,
+      "loss": 3.8815,
+      "step": 1783296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0793955959362223e-05,
+      "loss": 3.8919,
+      "step": 1783808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0785570011851703e-05,
+      "loss": 3.8899,
+      "step": 1784320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0777184064341183e-05,
+      "loss": 3.8897,
+      "step": 1784832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0768814495634395e-05,
+      "loss": 3.8972,
+      "step": 1785344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0760428548123875e-05,
+      "loss": 3.8891,
+      "step": 1785856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0752042600613355e-05,
+      "loss": 3.8845,
+      "step": 1786368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0743656653102832e-05,
+      "loss": 3.8985,
+      "step": 1786880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0735270705592312e-05,
+      "loss": 3.8788,
+      "step": 1787392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0726884758081795e-05,
+      "loss": 3.8891,
+      "step": 1787904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0718498810571275e-05,
+      "loss": 3.8934,
+      "step": 1788416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0710129241864484e-05,
+      "loss": 3.892,
+      "step": 1788928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0701743294353964e-05,
+      "loss": 3.8832,
+      "step": 1789440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0693357346843444e-05,
+      "loss": 3.8833,
+      "step": 1789952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0684987778136657e-05,
+      "loss": 3.8738,
+      "step": 1790464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0676601830626137e-05,
+      "loss": 3.892,
+      "step": 1790976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0668215883115616e-05,
+      "loss": 3.9026,
+      "step": 1791488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0659829935605096e-05,
+      "loss": 3.8927,
+      "step": 1792000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0651443988094576e-05,
+      "loss": 3.894,
+      "step": 1792512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0643058040584056e-05,
+      "loss": 3.8992,
+      "step": 1793024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0634672093073536e-05,
+      "loss": 3.9071,
+      "step": 1793536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0626286145563016e-05,
+      "loss": 3.8765,
+      "step": 1794048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0617900198052496e-05,
+      "loss": 3.897,
+      "step": 1794560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.060953062934571e-05,
+      "loss": 3.8974,
+      "step": 1795072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.060114468183519e-05,
+      "loss": 3.874,
+      "step": 1795584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0592758734324665e-05,
+      "loss": 3.9006,
+      "step": 1796096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0584372786814145e-05,
+      "loss": 3.8885,
+      "step": 1796608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0575986839303625e-05,
+      "loss": 3.9029,
+      "step": 1797120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.056760089179311e-05,
+      "loss": 3.8979,
+      "step": 1797632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.055921494428259e-05,
+      "loss": 3.8775,
+      "step": 1798144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0550845375575798e-05,
+      "loss": 3.8852,
+      "step": 1798656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0542459428065277e-05,
+      "loss": 3.8918,
+      "step": 1799168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0534073480554757e-05,
+      "loss": 3.8933,
+      "step": 1799680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0525687533044237e-05,
+      "loss": 3.8905,
+      "step": 1800192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0517301585533717e-05,
+      "loss": 3.8993,
+      "step": 1800704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0508915638023197e-05,
+      "loss": 3.8916,
+      "step": 1801216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0500529690512677e-05,
+      "loss": 3.8872,
+      "step": 1801728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0492143743002157e-05,
+      "loss": 3.8867,
+      "step": 1802240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.048377417429537e-05,
+      "loss": 3.8836,
+      "step": 1802752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.047540460558858e-05,
+      "loss": 3.8978,
+      "step": 1803264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.046701865807806e-05,
+      "loss": 3.8934,
+      "step": 1803776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0458632710567542e-05,
+      "loss": 3.9012,
+      "step": 1804288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.045024676305702e-05,
+      "loss": 3.8759,
+      "step": 1804800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.04418608155465e-05,
+      "loss": 3.8919,
+      "step": 1805312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.043347486803598e-05,
+      "loss": 3.8999,
+      "step": 1805824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.042510529932919e-05,
+      "loss": 3.8996,
+      "step": 1806336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.041671935181867e-05,
+      "loss": 3.8922,
+      "step": 1806848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.040833340430815e-05,
+      "loss": 3.8844,
+      "step": 1807360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.039994745679763e-05,
+      "loss": 3.8908,
+      "step": 1807872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0391577888090843e-05,
+      "loss": 3.8968,
+      "step": 1808384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.038319194058032e-05,
+      "loss": 3.8734,
+      "step": 1808896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.03748059930698e-05,
+      "loss": 3.891,
+      "step": 1809408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0366436424363012e-05,
+      "loss": 3.8775,
+      "step": 1809920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0358050476852492e-05,
+      "loss": 3.8967,
+      "step": 1810432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0349664529341972e-05,
+      "loss": 3.8851,
+      "step": 1810944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0341278581831452e-05,
+      "loss": 3.8943,
+      "step": 1811456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0332892634320932e-05,
+      "loss": 3.8888,
+      "step": 1811968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0324506686810412e-05,
+      "loss": 3.8958,
+      "step": 1812480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0316120739299892e-05,
+      "loss": 3.89,
+      "step": 1812992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0307734791789372e-05,
+      "loss": 3.8788,
+      "step": 1813504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0299348844278852e-05,
+      "loss": 3.8948,
+      "step": 1814016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0290962896768332e-05,
+      "loss": 3.8999,
+      "step": 1814528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0282593328061545e-05,
+      "loss": 3.8941,
+      "step": 1815040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0274223759354754e-05,
+      "loss": 3.8939,
+      "step": 1815552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0265837811844234e-05,
+      "loss": 3.881,
+      "step": 1816064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0257451864333717e-05,
+      "loss": 3.8961,
+      "step": 1816576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0249065916823197e-05,
+      "loss": 3.8943,
+      "step": 1817088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0240679969312673e-05,
+      "loss": 3.8853,
+      "step": 1817600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0232294021802153e-05,
+      "loss": 3.8845,
+      "step": 1818112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0223908074291633e-05,
+      "loss": 3.8906,
+      "step": 1818624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0215522126781113e-05,
+      "loss": 3.8931,
+      "step": 1819136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0207136179270593e-05,
+      "loss": 3.8922,
+      "step": 1819648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0198750231760073e-05,
+      "loss": 3.8733,
+      "step": 1820160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0190364284249557e-05,
+      "loss": 3.8959,
+      "step": 1820672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0181978336739033e-05,
+      "loss": 3.8806,
+      "step": 1821184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0173608768032246e-05,
+      "loss": 3.9028,
+      "step": 1821696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0165222820521726e-05,
+      "loss": 3.895,
+      "step": 1822208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0156836873011202e-05,
+      "loss": 3.8998,
+      "step": 1822720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0148450925500682e-05,
+      "loss": 3.8787,
+      "step": 1823232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0140064977990165e-05,
+      "loss": 3.9039,
+      "step": 1823744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0131679030479645e-05,
+      "loss": 3.8948,
+      "step": 1824256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0123309461772854e-05,
+      "loss": 3.8951,
+      "step": 1824768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0114923514262334e-05,
+      "loss": 3.888,
+      "step": 1825280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0106537566751814e-05,
+      "loss": 3.8928,
+      "step": 1825792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0098151619241294e-05,
+      "loss": 3.8968,
+      "step": 1826304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0089782050534507e-05,
+      "loss": 3.9063,
+      "step": 1826816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0081396103023987e-05,
+      "loss": 3.8847,
+      "step": 1827328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0073010155513467e-05,
+      "loss": 3.8941,
+      "step": 1827840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0064624208002947e-05,
+      "loss": 3.901,
+      "step": 1828352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0056254639296156e-05,
+      "loss": 3.8888,
+      "step": 1828864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.004788507058937e-05,
+      "loss": 3.8884,
+      "step": 1829376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.003949912307885e-05,
+      "loss": 3.8955,
+      "step": 1829888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0031113175568328e-05,
+      "loss": 3.8988,
+      "step": 1830400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0022727228057808e-05,
+      "loss": 3.8876,
+      "step": 1830912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0014341280547288e-05,
+      "loss": 3.8932,
+      "step": 1831424
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9896769523620605,
+      "eval_runtime": 298.823,
+      "eval_samples_per_second": 1276.98,
+      "eval_steps_per_second": 39.907,
+      "step": 1831680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0005955333036768e-05,
+      "loss": 3.8835,
+      "step": 1831936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.999758576432998e-05,
+      "loss": 3.8794,
+      "step": 1832448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.998919981681946e-05,
+      "loss": 3.8907,
+      "step": 1832960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.998081386930894e-05,
+      "loss": 3.8923,
+      "step": 1833472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.997242792179842e-05,
+      "loss": 3.902,
+      "step": 1833984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.99640419742879e-05,
+      "loss": 3.8913,
+      "step": 1834496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.995565602677738e-05,
+      "loss": 3.8877,
+      "step": 1835008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9947270079266857e-05,
+      "loss": 3.8865,
+      "step": 1835520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9938884131756337e-05,
+      "loss": 3.8866,
+      "step": 1836032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.993049818424582e-05,
+      "loss": 3.8925,
+      "step": 1836544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9922128615539033e-05,
+      "loss": 3.8964,
+      "step": 1837056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9913759046832242e-05,
+      "loss": 3.8978,
+      "step": 1837568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.990538947812545e-05,
+      "loss": 3.8874,
+      "step": 1838080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9897003530614934e-05,
+      "loss": 3.8893,
+      "step": 1838592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9888617583104414e-05,
+      "loss": 3.8893,
+      "step": 1839104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9880231635593894e-05,
+      "loss": 3.8714,
+      "step": 1839616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9871845688083374e-05,
+      "loss": 3.8859,
+      "step": 1840128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9863459740572854e-05,
+      "loss": 3.8849,
+      "step": 1840640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.985507379306233e-05,
+      "loss": 3.888,
+      "step": 1841152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.984668784555181e-05,
+      "loss": 3.8942,
+      "step": 1841664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.983830189804129e-05,
+      "loss": 3.9025,
+      "step": 1842176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9829915950530774e-05,
+      "loss": 3.8925,
+      "step": 1842688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9821530003020254e-05,
+      "loss": 3.894,
+      "step": 1843200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9813144055509734e-05,
+      "loss": 3.8961,
+      "step": 1843712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9804774486802943e-05,
+      "loss": 3.8854,
+      "step": 1844224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9796404918096155e-05,
+      "loss": 3.8909,
+      "step": 1844736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9788018970585635e-05,
+      "loss": 3.89,
+      "step": 1845248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9779633023075115e-05,
+      "loss": 3.888,
+      "step": 1845760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9771247075564595e-05,
+      "loss": 3.8847,
+      "step": 1846272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9762861128054075e-05,
+      "loss": 3.8787,
+      "step": 1846784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9754475180543555e-05,
+      "loss": 3.8881,
+      "step": 1847296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9746089233033035e-05,
+      "loss": 3.8995,
+      "step": 1847808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.973770328552251e-05,
+      "loss": 3.8934,
+      "step": 1848320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9729333716815728e-05,
+      "loss": 3.8942,
+      "step": 1848832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9720947769305207e-05,
+      "loss": 3.8879,
+      "step": 1849344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9712578200598417e-05,
+      "loss": 3.898,
+      "step": 1849856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9704192253087897e-05,
+      "loss": 3.8922,
+      "step": 1850368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9695806305577376e-05,
+      "loss": 3.8817,
+      "step": 1850880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9687420358066856e-05,
+      "loss": 3.8811,
+      "step": 1851392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9679034410556336e-05,
+      "loss": 3.8798,
+      "step": 1851904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9670648463045816e-05,
+      "loss": 3.8807,
+      "step": 1852416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9662262515535296e-05,
+      "loss": 3.9013,
+      "step": 1852928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.965389294682851e-05,
+      "loss": 3.8897,
+      "step": 1853440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9645523378121718e-05,
+      "loss": 3.8868,
+      "step": 1853952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9637137430611198e-05,
+      "loss": 3.897,
+      "step": 1854464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.962875148310068e-05,
+      "loss": 3.8873,
+      "step": 1854976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.962036553559016e-05,
+      "loss": 3.8983,
+      "step": 1855488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9611979588079638e-05,
+      "loss": 3.8861,
+      "step": 1856000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.960361001937285e-05,
+      "loss": 3.8649,
+      "step": 1856512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.959522407186233e-05,
+      "loss": 3.9059,
+      "step": 1857024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.958683812435181e-05,
+      "loss": 3.8892,
+      "step": 1857536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.957845217684129e-05,
+      "loss": 3.8883,
+      "step": 1858048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.957006622933077e-05,
+      "loss": 3.8813,
+      "step": 1858560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9561696660623982e-05,
+      "loss": 3.8803,
+      "step": 1859072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.955331071311346e-05,
+      "loss": 3.88,
+      "step": 1859584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.954492476560294e-05,
+      "loss": 3.8855,
+      "step": 1860096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.953655519689615e-05,
+      "loss": 3.8831,
+      "step": 1860608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9528169249385635e-05,
+      "loss": 3.882,
+      "step": 1861120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951978330187511e-05,
+      "loss": 3.8934,
+      "step": 1861632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951139735436459e-05,
+      "loss": 3.8809,
+      "step": 1862144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.950301140685407e-05,
+      "loss": 3.8752,
+      "step": 1862656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.949462545934355e-05,
+      "loss": 3.8949,
+      "step": 1863168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.948623951183303e-05,
+      "loss": 3.8711,
+      "step": 1863680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.947785356432251e-05,
+      "loss": 3.8773,
+      "step": 1864192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.946946761681199e-05,
+      "loss": 3.8929,
+      "step": 1864704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.946108166930147e-05,
+      "loss": 3.8871,
+      "step": 1865216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.945269572179095e-05,
+      "loss": 3.8759,
+      "step": 1865728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.944430977428043e-05,
+      "loss": 3.8817,
+      "step": 1866240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.943594020557364e-05,
+      "loss": 3.8706,
+      "step": 1866752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9427570636866853e-05,
+      "loss": 3.8846,
+      "step": 1867264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9419184689356336e-05,
+      "loss": 3.8927,
+      "step": 1867776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9410798741845816e-05,
+      "loss": 3.8853,
+      "step": 1868288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9402412794335292e-05,
+      "loss": 3.8873,
+      "step": 1868800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9394026846824772e-05,
+      "loss": 3.8965,
+      "step": 1869312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9385640899314252e-05,
+      "loss": 3.8991,
+      "step": 1869824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9377271330607465e-05,
+      "loss": 3.8715,
+      "step": 1870336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9368885383096945e-05,
+      "loss": 3.8935,
+      "step": 1870848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9360499435586425e-05,
+      "loss": 3.8864,
+      "step": 1871360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9352113488075905e-05,
+      "loss": 3.8693,
+      "step": 1871872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9343727540565385e-05,
+      "loss": 3.8898,
+      "step": 1872384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9335357971858594e-05,
+      "loss": 3.8851,
+      "step": 1872896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9326972024348074e-05,
+      "loss": 3.897,
+      "step": 1873408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9318586076837557e-05,
+      "loss": 3.8893,
+      "step": 1873920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9310200129327037e-05,
+      "loss": 3.8753,
+      "step": 1874432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9301814181816517e-05,
+      "loss": 3.8772,
+      "step": 1874944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9293428234305997e-05,
+      "loss": 3.8877,
+      "step": 1875456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9285042286795474e-05,
+      "loss": 3.8856,
+      "step": 1875968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9276656339284953e-05,
+      "loss": 3.8903,
+      "step": 1876480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9268286770578166e-05,
+      "loss": 3.889,
+      "step": 1876992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9259900823067646e-05,
+      "loss": 3.8842,
+      "step": 1877504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9251514875557126e-05,
+      "loss": 3.8866,
+      "step": 1878016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9243128928046606e-05,
+      "loss": 3.8795,
+      "step": 1878528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9234759359339818e-05,
+      "loss": 3.8774,
+      "step": 1879040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9226373411829295e-05,
+      "loss": 3.8935,
+      "step": 1879552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9217987464318775e-05,
+      "loss": 3.8853,
+      "step": 1880064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9209601516808258e-05,
+      "loss": 3.8948,
+      "step": 1880576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.920123194810147e-05,
+      "loss": 3.8713,
+      "step": 1881088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9192846000590947e-05,
+      "loss": 3.8862,
+      "step": 1881600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9184460053080427e-05,
+      "loss": 3.8928,
+      "step": 1882112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9176074105569907e-05,
+      "loss": 3.8893,
+      "step": 1882624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.916770453686312e-05,
+      "loss": 3.8852,
+      "step": 1883136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.91593185893526e-05,
+      "loss": 3.8802,
+      "step": 1883648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.915093264184208e-05,
+      "loss": 3.8854,
+      "step": 1884160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.914254669433156e-05,
+      "loss": 3.8865,
+      "step": 1884672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.913417712562477e-05,
+      "loss": 3.8731,
+      "step": 1885184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.912579117811425e-05,
+      "loss": 3.8804,
+      "step": 1885696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.911742160940746e-05,
+      "loss": 3.8737,
+      "step": 1886208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.910903566189694e-05,
+      "loss": 3.8933,
+      "step": 1886720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.910064971438642e-05,
+      "loss": 3.8773,
+      "step": 1887232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.90922637668759e-05,
+      "loss": 3.8872,
+      "step": 1887744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.908387781936538e-05,
+      "loss": 3.8797,
+      "step": 1888256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.907549187185486e-05,
+      "loss": 3.8942,
+      "step": 1888768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.906710592434434e-05,
+      "loss": 3.8802,
+      "step": 1889280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.905871997683382e-05,
+      "loss": 3.8737,
+      "step": 1889792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9050350408127033e-05,
+      "loss": 3.8915,
+      "step": 1890304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9041964460616513e-05,
+      "loss": 3.8933,
+      "step": 1890816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9033578513105993e-05,
+      "loss": 3.8892,
+      "step": 1891328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9025208944399202e-05,
+      "loss": 3.8866,
+      "step": 1891840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9016822996888682e-05,
+      "loss": 3.8758,
+      "step": 1892352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9008437049378166e-05,
+      "loss": 3.8884,
+      "step": 1892864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9000051101867645e-05,
+      "loss": 3.8882,
+      "step": 1893376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8991665154357122e-05,
+      "loss": 3.8802,
+      "step": 1893888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8983279206846602e-05,
+      "loss": 3.8756,
+      "step": 1894400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8974893259336082e-05,
+      "loss": 3.8851,
+      "step": 1894912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8966507311825562e-05,
+      "loss": 3.8891,
+      "step": 1895424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8958121364315042e-05,
+      "loss": 3.8865,
+      "step": 1895936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8949751795608254e-05,
+      "loss": 3.8685,
+      "step": 1896448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8941365848097734e-05,
+      "loss": 3.888,
+      "step": 1896960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8932979900587214e-05,
+      "loss": 3.8759,
+      "step": 1897472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8924593953076694e-05,
+      "loss": 3.8939,
+      "step": 1897984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8916208005566174e-05,
+      "loss": 3.8884,
+      "step": 1898496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8907838436859383e-05,
+      "loss": 3.8907,
+      "step": 1899008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8899452489348867e-05,
+      "loss": 3.8715,
+      "step": 1899520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8891082920642076e-05,
+      "loss": 3.8985,
+      "step": 1900032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8882696973131556e-05,
+      "loss": 3.888,
+      "step": 1900544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8874327404424768e-05,
+      "loss": 3.8918,
+      "step": 1901056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8865941456914245e-05,
+      "loss": 3.8839,
+      "step": 1901568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8857555509403728e-05,
+      "loss": 3.8842,
+      "step": 1902080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8849169561893208e-05,
+      "loss": 3.8893,
+      "step": 1902592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8840783614382688e-05,
+      "loss": 3.9046,
+      "step": 1903104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8832397666872168e-05,
+      "loss": 3.879,
+      "step": 1903616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8824011719361648e-05,
+      "loss": 3.8882,
+      "step": 1904128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8815625771851128e-05,
+      "loss": 3.8937,
+      "step": 1904640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8807239824340604e-05,
+      "loss": 3.8794,
+      "step": 1905152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.879887025563382e-05,
+      "loss": 3.8823,
+      "step": 1905664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.87904843081233e-05,
+      "loss": 3.8878,
+      "step": 1906176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.878211473941651e-05,
+      "loss": 3.8948,
+      "step": 1906688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.877372879190599e-05,
+      "loss": 3.8858,
+      "step": 1907200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.876534284439547e-05,
+      "loss": 3.8843,
+      "step": 1907712
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.987194299697876,
+      "eval_runtime": 300.0644,
+      "eval_samples_per_second": 1271.697,
+      "eval_steps_per_second": 39.741,
+      "step": 1908000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.875695689688495e-05,
+      "loss": 3.8766,
+      "step": 1908224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.874857094937443e-05,
+      "loss": 3.8731,
+      "step": 1908736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.874018500186391e-05,
+      "loss": 3.8861,
+      "step": 1909248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.873179905435339e-05,
+      "loss": 3.8887,
+      "step": 1909760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.872341310684287e-05,
+      "loss": 3.894,
+      "step": 1910272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.871502715933235e-05,
+      "loss": 3.8829,
+      "step": 1910784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.870664121182183e-05,
+      "loss": 3.8856,
+      "step": 1911296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.869825526431131e-05,
+      "loss": 3.8784,
+      "step": 1911808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8689869316800785e-05,
+      "loss": 3.8793,
+      "step": 1912320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.868148336929027e-05,
+      "loss": 3.8876,
+      "step": 1912832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.867309742177975e-05,
+      "loss": 3.8911,
+      "step": 1913344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.866471147426923e-05,
+      "loss": 3.8889,
+      "step": 1913856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8656341905562438e-05,
+      "loss": 3.8905,
+      "step": 1914368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8647955958051918e-05,
+      "loss": 3.8771,
+      "step": 1914880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8639570010541398e-05,
+      "loss": 3.8795,
+      "step": 1915392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8631184063030878e-05,
+      "loss": 3.8685,
+      "step": 1915904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.862279811552036e-05,
+      "loss": 3.8796,
+      "step": 1916416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.861441216800984e-05,
+      "loss": 3.8818,
+      "step": 1916928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8606026220499318e-05,
+      "loss": 3.8802,
+      "step": 1917440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8597640272988797e-05,
+      "loss": 3.8866,
+      "step": 1917952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8589254325478277e-05,
+      "loss": 3.897,
+      "step": 1918464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.858088475677149e-05,
+      "loss": 3.8881,
+      "step": 1918976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.857249880926097e-05,
+      "loss": 3.8885,
+      "step": 1919488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.856411286175045e-05,
+      "loss": 3.8864,
+      "step": 1920000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.855572691423993e-05,
+      "loss": 3.8795,
+      "step": 1920512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.854737372433687e-05,
+      "loss": 3.8857,
+      "step": 1921024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.853898777682635e-05,
+      "loss": 3.8843,
+      "step": 1921536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.853060182931583e-05,
+      "loss": 3.8834,
+      "step": 1922048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8522215881805315e-05,
+      "loss": 3.8761,
+      "step": 1922560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.851382993429479e-05,
+      "loss": 3.8758,
+      "step": 1923072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.850544398678427e-05,
+      "loss": 3.878,
+      "step": 1923584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.849705803927375e-05,
+      "loss": 3.8926,
+      "step": 1924096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.848867209176323e-05,
+      "loss": 3.8896,
+      "step": 1924608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.848028614425271e-05,
+      "loss": 3.888,
+      "step": 1925120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.847190019674219e-05,
+      "loss": 3.8786,
+      "step": 1925632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8463547006839136e-05,
+      "loss": 3.8879,
+      "step": 1926144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8455161059328613e-05,
+      "loss": 3.8902,
+      "step": 1926656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8446775111818093e-05,
+      "loss": 3.8768,
+      "step": 1927168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8438389164307573e-05,
+      "loss": 3.8768,
+      "step": 1927680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8430003216797052e-05,
+      "loss": 3.8709,
+      "step": 1928192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8421617269286532e-05,
+      "loss": 3.8787,
+      "step": 1928704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8413231321776016e-05,
+      "loss": 3.8916,
+      "step": 1929216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8404845374265496e-05,
+      "loss": 3.8872,
+      "step": 1929728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8396475805558705e-05,
+      "loss": 3.8799,
+      "step": 1930240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8388089858048185e-05,
+      "loss": 3.8897,
+      "step": 1930752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8379703910537665e-05,
+      "loss": 3.8804,
+      "step": 1931264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8371317963027145e-05,
+      "loss": 3.888,
+      "step": 1931776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8362932015516625e-05,
+      "loss": 3.8871,
+      "step": 1932288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8354546068006105e-05,
+      "loss": 3.8562,
+      "step": 1932800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8346176499299317e-05,
+      "loss": 3.9031,
+      "step": 1933312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8337790551788794e-05,
+      "loss": 3.8795,
+      "step": 1933824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8329404604278274e-05,
+      "loss": 3.8921,
+      "step": 1934336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8321018656767754e-05,
+      "loss": 3.872,
+      "step": 1934848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8312632709257237e-05,
+      "loss": 3.8783,
+      "step": 1935360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8304246761746717e-05,
+      "loss": 3.8723,
+      "step": 1935872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8295860814236197e-05,
+      "loss": 3.8786,
+      "step": 1936384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8287474866725677e-05,
+      "loss": 3.8771,
+      "step": 1936896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8279105298018886e-05,
+      "loss": 3.8742,
+      "step": 1937408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8270719350508366e-05,
+      "loss": 3.8921,
+      "step": 1937920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8262333402997846e-05,
+      "loss": 3.8749,
+      "step": 1938432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8253947455487326e-05,
+      "loss": 3.8684,
+      "step": 1938944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8245577886780538e-05,
+      "loss": 3.889,
+      "step": 1939456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8237191939270018e-05,
+      "loss": 3.8652,
+      "step": 1939968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8228822370563227e-05,
+      "loss": 3.8716,
+      "step": 1940480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8220436423052707e-05,
+      "loss": 3.8869,
+      "step": 1940992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.821205047554219e-05,
+      "loss": 3.8799,
+      "step": 1941504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.820366452803167e-05,
+      "loss": 3.8727,
+      "step": 1942016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.819527858052115e-05,
+      "loss": 3.8701,
+      "step": 1942528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.818690901181436e-05,
+      "loss": 3.8692,
+      "step": 1943040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.817852306430384e-05,
+      "loss": 3.8782,
+      "step": 1943552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.817013711679332e-05,
+      "loss": 3.8891,
+      "step": 1944064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.81617511692828e-05,
+      "loss": 3.8792,
+      "step": 1944576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.815336522177228e-05,
+      "loss": 3.8835,
+      "step": 1945088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.814497927426176e-05,
+      "loss": 3.8905,
+      "step": 1945600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.813659332675124e-05,
+      "loss": 3.8968,
+      "step": 1946112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.812820737924072e-05,
+      "loss": 3.8646,
+      "step": 1946624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.811983781053393e-05,
+      "loss": 3.8903,
+      "step": 1947136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.811145186302341e-05,
+      "loss": 3.8775,
+      "step": 1947648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.810306591551289e-05,
+      "loss": 3.8674,
+      "step": 1948160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.809467996800237e-05,
+      "loss": 3.8816,
+      "step": 1948672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.808631039929558e-05,
+      "loss": 3.8876,
+      "step": 1949184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.807792445178506e-05,
+      "loss": 3.8884,
+      "step": 1949696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.806953850427454e-05,
+      "loss": 3.8906,
+      "step": 1950208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.806115255676402e-05,
+      "loss": 3.8662,
+      "step": 1950720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.80527666092535e-05,
+      "loss": 3.8693,
+      "step": 1951232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.804438066174298e-05,
+      "loss": 3.8845,
+      "step": 1951744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.803599471423246e-05,
+      "loss": 3.8786,
+      "step": 1952256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.802760876672194e-05,
+      "loss": 3.8811,
+      "step": 1952768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8019255576818882e-05,
+      "loss": 3.888,
+      "step": 1953280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8010869629308362e-05,
+      "loss": 3.8792,
+      "step": 1953792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8002483681797845e-05,
+      "loss": 3.8782,
+      "step": 1954304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7994097734287325e-05,
+      "loss": 3.8769,
+      "step": 1954816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7985728165580534e-05,
+      "loss": 3.8717,
+      "step": 1955328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7977342218070014e-05,
+      "loss": 3.888,
+      "step": 1955840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7968956270559494e-05,
+      "loss": 3.8827,
+      "step": 1956352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7960570323048974e-05,
+      "loss": 3.8839,
+      "step": 1956864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7952184375538454e-05,
+      "loss": 3.8717,
+      "step": 1957376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7943798428027934e-05,
+      "loss": 3.8765,
+      "step": 1957888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7935412480517414e-05,
+      "loss": 3.8883,
+      "step": 1958400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7927026533006894e-05,
+      "loss": 3.8844,
+      "step": 1958912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7918640585496374e-05,
+      "loss": 3.8827,
+      "step": 1959424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7910271016789583e-05,
+      "loss": 3.878,
+      "step": 1959936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7901885069279063e-05,
+      "loss": 3.8774,
+      "step": 1960448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7893499121768546e-05,
+      "loss": 3.8759,
+      "step": 1960960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7885113174258026e-05,
+      "loss": 3.8743,
+      "step": 1961472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7876743605551235e-05,
+      "loss": 3.8701,
+      "step": 1961984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7868357658040715e-05,
+      "loss": 3.8673,
+      "step": 1962496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7859971710530195e-05,
+      "loss": 3.8888,
+      "step": 1963008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7851585763019675e-05,
+      "loss": 3.8748,
+      "step": 1963520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7843216194312888e-05,
+      "loss": 3.8797,
+      "step": 1964032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7834830246802368e-05,
+      "loss": 3.8704,
+      "step": 1964544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7826444299291848e-05,
+      "loss": 3.8879,
+      "step": 1965056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7818058351781328e-05,
+      "loss": 3.8775,
+      "step": 1965568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7809688783074537e-05,
+      "loss": 3.8702,
+      "step": 1966080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7801302835564017e-05,
+      "loss": 3.8887,
+      "step": 1966592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.779293326685723e-05,
+      "loss": 3.887,
+      "step": 1967104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.778454731934671e-05,
+      "loss": 3.8841,
+      "step": 1967616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.777616137183619e-05,
+      "loss": 3.879,
+      "step": 1968128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.776777542432567e-05,
+      "loss": 3.874,
+      "step": 1968640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.775938947681515e-05,
+      "loss": 3.8798,
+      "step": 1969152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.775100352930463e-05,
+      "loss": 3.8838,
+      "step": 1969664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.774261758179411e-05,
+      "loss": 3.8728,
+      "step": 1970176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.773423163428359e-05,
+      "loss": 3.874,
+      "step": 1970688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.77258620655768e-05,
+      "loss": 3.8809,
+      "step": 1971200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.771747611806628e-05,
+      "loss": 3.8831,
+      "step": 1971712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7709090170555758e-05,
+      "loss": 3.88,
+      "step": 1972224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7700704223045238e-05,
+      "loss": 3.8655,
+      "step": 1972736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7692334654338454e-05,
+      "loss": 3.884,
+      "step": 1973248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7683965085631663e-05,
+      "loss": 3.8686,
+      "step": 1973760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7675579138121143e-05,
+      "loss": 3.8899,
+      "step": 1974272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7667193190610623e-05,
+      "loss": 3.8823,
+      "step": 1974784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7658807243100103e-05,
+      "loss": 3.8913,
+      "step": 1975296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.765042129558958e-05,
+      "loss": 3.8634,
+      "step": 1975808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7642035348079063e-05,
+      "loss": 3.8949,
+      "step": 1976320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7633649400568543e-05,
+      "loss": 3.8791,
+      "step": 1976832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7625263453058023e-05,
+      "loss": 3.8895,
+      "step": 1977344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.761689388435123e-05,
+      "loss": 3.8811,
+      "step": 1977856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.760850793684071e-05,
+      "loss": 3.8729,
+      "step": 1978368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.760012198933019e-05,
+      "loss": 3.8896,
+      "step": 1978880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.759173604181967e-05,
+      "loss": 3.8958,
+      "step": 1979392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7583366473112884e-05,
+      "loss": 3.8792,
+      "step": 1979904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7574980525602364e-05,
+      "loss": 3.879,
+      "step": 1980416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7566594578091844e-05,
+      "loss": 3.8885,
+      "step": 1980928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7558208630581324e-05,
+      "loss": 3.8749,
+      "step": 1981440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7549839061874533e-05,
+      "loss": 3.8794,
+      "step": 1981952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7541453114364016e-05,
+      "loss": 3.8862,
+      "step": 1982464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7533067166853496e-05,
+      "loss": 3.8842,
+      "step": 1982976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7524681219342976e-05,
+      "loss": 3.8842,
+      "step": 1983488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7516311650636185e-05,
+      "loss": 3.8776,
+      "step": 1984000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.985201120376587,
+      "eval_runtime": 295.7878,
+      "eval_samples_per_second": 1290.084,
+      "eval_steps_per_second": 40.316,
+      "step": 1984320
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 8.177127209909868e+17,
+  "trial_name": null,
+  "trial_params": null
+}