diff --git "a/checkpoint-1373760/trainer_state.json" "b/checkpoint-1373760/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1373760/trainer_state.json"
@@ -0,0 +1,16267 @@
+{
+  "best_metric": 4.026576519012451,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/lstm/3/checkpoints/checkpoint-1373760",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1373760,
+  "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": 10.8201,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5492,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0526,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9871,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9407,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8843,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7378,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6415,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5324,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4537,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3806,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3206,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.252,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1814,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1257,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0627,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0237,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 5.9827,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9414,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9069,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.879,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8403,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8174,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7748,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7628,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7318,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.712,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6865,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6532,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6376,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 5.6133,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 5.6036,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5765,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5648,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5546,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5318,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5215,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.4777,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4722,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4533,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.439,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.434,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4257,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.3975,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3965,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3909,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.379,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3733,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3563,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3349,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3376,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3174,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3197,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2887,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.3009,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2731,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2915,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2634,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2499,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2446,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2396,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2252,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2266,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.198,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.195,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2061,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1981,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1816,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1626,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1623,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1629,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1562,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1519,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1479,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1494,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.1399,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1137,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1105,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.107,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.095,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1048,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0855,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.1028,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0783,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0619,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0673,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0676,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0641,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0502,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0465,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0385,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 5.0372,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.025,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0226,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.021,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0256,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.0058,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9998,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9933,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.9894,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 4.9939,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9811,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9736,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9654,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9688,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9409,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.956,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9429,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9493,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9367,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9451,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9312,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9244,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.9322,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9104,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9269,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9158,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9131,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.9006,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.9007,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9025,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8934,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8928,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8821,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8854,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8653,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8692,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8529,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8585,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8461,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8661,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.8716,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8553,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8336,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8458,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8442,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.838,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8328,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8336,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8306,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8187,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.8199,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.819,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8069,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8075,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8154,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.8038,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8018,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8134,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.7907,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.779759407043457,
+      "eval_runtime": 295.6929,
+      "eval_samples_per_second": 1290.498,
+      "eval_steps_per_second": 40.329,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.7784,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.784,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7984,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7851,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7911,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.765,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7833,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7554,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.7691,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.7587,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658690626017536e-05,
+      "loss": 4.7718,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650304678507016e-05,
+      "loss": 4.7685,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7562,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7493,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.736,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.7379,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.748,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7439,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7368,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7517,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7459,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7374,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7243,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.7455,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854133649728145e-05,
+      "loss": 4.7214,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853295054977093e-05,
+      "loss": 4.731,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.7272,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.7206,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85078090860431e-05,
+      "loss": 4.7096,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849942313853258e-05,
+      "loss": 4.7089,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849103719102206e-05,
+      "loss": 4.7119,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848265124351154e-05,
+      "loss": 4.7087,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847426529600102e-05,
+      "loss": 4.7094,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84658793484905e-05,
+      "loss": 4.7141,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.7046,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449123832273194e-05,
+      "loss": 4.7035,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440737884762674e-05,
+      "loss": 4.6828,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432351937252154e-05,
+      "loss": 4.6883,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423965989741634e-05,
+      "loss": 4.6799,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415580042231114e-05,
+      "loss": 4.677,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840719409472059e-05,
+      "loss": 4.6866,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83988245260138e-05,
+      "loss": 4.6895,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839043857850328e-05,
+      "loss": 4.6722,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838205263099276e-05,
+      "loss": 4.679,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373666683482236e-05,
+      "loss": 4.6768,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836529711477545e-05,
+      "loss": 4.6786,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835691116726493e-05,
+      "loss": 4.6803,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834852521975441e-05,
+      "loss": 4.6717,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834013927224389e-05,
+      "loss": 4.6604,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833175332473337e-05,
+      "loss": 4.6692,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832338375602659e-05,
+      "loss": 4.6563,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831499780851606e-05,
+      "loss": 4.6655,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830661186100554e-05,
+      "loss": 4.6524,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829822591349502e-05,
+      "loss": 4.663,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6406,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828147039727771e-05,
+      "loss": 4.6642,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827308444976719e-05,
+      "loss": 4.6484,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826469850225667e-05,
+      "loss": 4.6416,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.6506,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247942986039366e-05,
+      "loss": 4.6417,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239557038528846e-05,
+      "loss": 4.6337,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6483,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6179,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6246,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.6365,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197643679779975e-05,
+      "loss": 4.6366,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189257732269455e-05,
+      "loss": 4.6268,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180871784758935e-05,
+      "loss": 4.6177,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172485837248415e-05,
+      "loss": 4.6172,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164116268541624e-05,
+      "loss": 4.6221,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155730321031104e-05,
+      "loss": 4.6273,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147344373520584e-05,
+      "loss": 4.6222,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.623,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813057247849955e-05,
+      "loss": 4.628,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812218653098903e-05,
+      "loss": 4.6244,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.6059,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6068,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.6139,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808865911975068e-05,
+      "loss": 4.5986,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808027317224016e-05,
+      "loss": 4.6103,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807188722472964e-05,
+      "loss": 4.5963,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806350127721912e-05,
+      "loss": 4.6213,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80551153297086e-05,
+      "loss": 4.6016,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804674576100181e-05,
+      "loss": 4.5852,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803835981349129e-05,
+      "loss": 4.5957,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802997386598077e-05,
+      "loss": 4.5981,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802158791847025e-05,
+      "loss": 4.6022,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801320197095973e-05,
+      "loss": 4.5909,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004832402252944e-05,
+      "loss": 4.5929,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996446454742424e-05,
+      "loss": 4.5963,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79880605072319e-05,
+      "loss": 4.5843,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797967455972138e-05,
+      "loss": 4.5836,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797128861221086e-05,
+      "loss": 4.582,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796291904350407e-05,
+      "loss": 4.59,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954533095993546e-05,
+      "loss": 4.5858,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946147148483026e-05,
+      "loss": 4.5826,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937761200972506e-05,
+      "loss": 4.5722,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929375253461986e-05,
+      "loss": 4.5716,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792098930595147e-05,
+      "loss": 4.5769,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791260335844095e-05,
+      "loss": 4.5757,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790421741093043e-05,
+      "loss": 4.5711,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789584784222364e-05,
+      "loss": 4.5677,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788747827351685e-05,
+      "loss": 4.5597,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787909232600633e-05,
+      "loss": 4.5716,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787070637849581e-05,
+      "loss": 4.5405,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786232043098529e-05,
+      "loss": 4.5647,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785393448347477e-05,
+      "loss": 4.5553,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784554853596425e-05,
+      "loss": 4.5642,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783716258845373e-05,
+      "loss": 4.551,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782879301974694e-05,
+      "loss": 4.5646,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782040707223642e-05,
+      "loss": 4.5518,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812021124725906e-05,
+      "loss": 4.5493,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803635177215386e-05,
+      "loss": 4.5595,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795249229704866e-05,
+      "loss": 4.5427,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786863282194346e-05,
+      "loss": 4.5573,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778477334683826e-05,
+      "loss": 4.5556,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770107765977035e-05,
+      "loss": 4.5497,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761721818466515e-05,
+      "loss": 4.5415,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753335870955995e-05,
+      "loss": 4.5465,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7744966302249204e-05,
+      "loss": 4.5487,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736580354738684e-05,
+      "loss": 4.5352,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728194407228164e-05,
+      "loss": 4.551,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719808459717644e-05,
+      "loss": 4.5335,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771143889101086e-05,
+      "loss": 4.5419,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770305294350034e-05,
+      "loss": 4.5344,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769466699598982e-05,
+      "loss": 4.532,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76862810484793e-05,
+      "loss": 4.5148,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767791147977251e-05,
+      "loss": 4.5305,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766952553226199e-05,
+      "loss": 4.5166,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766113958475147e-05,
+      "loss": 4.5387,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765275363724095e-05,
+      "loss": 4.5492,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764436768973043e-05,
+      "loss": 4.5322,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76359817422199e-05,
+      "loss": 4.515,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762759579470938e-05,
+      "loss": 4.5288,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761920984719886e-05,
+      "loss": 4.5234,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761082389968834e-05,
+      "loss": 4.5265,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760243795217783e-05,
+      "loss": 4.5231,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759405200466731e-05,
+      "loss": 4.5219,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758566605715679e-05,
+      "loss": 4.5186,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757731286725373e-05,
+      "loss": 4.5164,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7568926919743206e-05,
+      "loss": 4.5163,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560540972232686e-05,
+      "loss": 4.521,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552155024722166e-05,
+      "loss": 4.5128,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543785456015375e-05,
+      "loss": 4.5066,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535399508504855e-05,
+      "loss": 4.5203,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527013560994335e-05,
+      "loss": 4.5137,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.5045,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510241665973295e-05,
+      "loss": 4.5231,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750185571846278e-05,
+      "loss": 4.5038,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.4971795082092285,
+      "eval_runtime": 294.126,
+      "eval_samples_per_second": 1297.373,
+      "eval_steps_per_second": 40.544,
+      "step": 152640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.749346977095226e-05,
+      "loss": 4.4931,
+      "step": 153088
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.748508382344174e-05,
+      "loss": 4.4975,
+      "step": 153600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.747671425473495e-05,
+      "loss": 4.5146,
+      "step": 154112
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.746834468602816e-05,
+      "loss": 4.505,
+      "step": 154624
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.745995873851764e-05,
+      "loss": 4.5102,
+      "step": 155136
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.745158916981085e-05,
+      "loss": 4.4876,
+      "step": 155648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.744320322230033e-05,
+      "loss": 4.5064,
+      "step": 156160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.743481727478981e-05,
+      "loss": 4.4791,
+      "step": 156672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.742643132727929e-05,
+      "loss": 4.5046,
+      "step": 157184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.741804537976877e-05,
+      "loss": 4.4865,
+      "step": 157696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.740965943225825e-05,
+      "loss": 4.5028,
+      "step": 158208
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7401273484747736e-05,
+      "loss": 4.5012,
+      "step": 158720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7392903916040945e-05,
+      "loss": 4.4898,
+      "step": 159232
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7384517968530425e-05,
+      "loss": 4.4859,
+      "step": 159744
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7376132021019905e-05,
+      "loss": 4.4796,
+      "step": 160256
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7367746073509385e-05,
+      "loss": 4.4778,
+      "step": 160768
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7359360125998865e-05,
+      "loss": 4.4841,
+      "step": 161280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7350974178488345e-05,
+      "loss": 4.4842,
+      "step": 161792
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7342604609781554e-05,
+      "loss": 4.484,
+      "step": 162304
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7334218662271034e-05,
+      "loss": 4.4952,
+      "step": 162816
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7325832714760514e-05,
+      "loss": 4.4936,
+      "step": 163328
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7317446767249994e-05,
+      "loss": 4.4881,
+      "step": 163840
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7309060819739474e-05,
+      "loss": 4.4777,
+      "step": 164352
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7300674872228953e-05,
+      "loss": 4.4904,
+      "step": 164864
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7292288924718433e-05,
+      "loss": 4.476,
+      "step": 165376
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.728390297720791e-05,
+      "loss": 4.4827,
+      "step": 165888
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.727551702969739e-05,
+      "loss": 4.4823,
+      "step": 166400
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.726713108218687e-05,
+      "loss": 4.4769,
+      "step": 166912
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.725874513467635e-05,
+      "loss": 4.4666,
+      "step": 167424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.725037556596957e-05,
+      "loss": 4.4681,
+      "step": 167936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.724198961845904e-05,
+      "loss": 4.4715,
+      "step": 168448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.723360367094852e-05,
+      "loss": 4.4733,
+      "step": 168960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7225217723438e-05,
+      "loss": 4.4762,
+      "step": 169472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.721683177592748e-05,
+      "loss": 4.4726,
+      "step": 169984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.4689,
+      "step": 170496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.4692,
+      "step": 171008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4526,
+      "step": 171520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.718330436468914e-05,
+      "loss": 4.4615,
+      "step": 172032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.717491841717862e-05,
+      "loss": 4.4528,
+      "step": 172544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.71665324696681e-05,
+      "loss": 4.4521,
+      "step": 173056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.715816290096131e-05,
+      "loss": 4.4575,
+      "step": 173568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.714977695345079e-05,
+      "loss": 4.4652,
+      "step": 174080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.714139100594027e-05,
+      "loss": 4.4516,
+      "step": 174592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.713300505842975e-05,
+      "loss": 4.4523,
+      "step": 175104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7124635489722956e-05,
+      "loss": 4.4615,
+      "step": 175616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7116249542212436e-05,
+      "loss": 4.4527,
+      "step": 176128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7107863594701916e-05,
+      "loss": 4.4667,
+      "step": 176640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7099477647191396e-05,
+      "loss": 4.4512,
+      "step": 177152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.7091108078484605e-05,
+      "loss": 4.4405,
+      "step": 177664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.708272213097409e-05,
+      "loss": 4.4557,
+      "step": 178176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.707433618346357e-05,
+      "loss": 4.4432,
+      "step": 178688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.706595023595305e-05,
+      "loss": 4.4522,
+      "step": 179200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.705756428844253e-05,
+      "loss": 4.4412,
+      "step": 179712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.704917834093201e-05,
+      "loss": 4.4469,
+      "step": 180224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.704079239342149e-05,
+      "loss": 4.4371,
+      "step": 180736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.70324228247147e-05,
+      "loss": 4.4486,
+      "step": 181248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.702403687720418e-05,
+      "loss": 4.4392,
+      "step": 181760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.701565092969366e-05,
+      "loss": 4.4362,
+      "step": 182272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.700726498218314e-05,
+      "loss": 4.4442,
+      "step": 182784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.699889541347635e-05,
+      "loss": 4.4328,
+      "step": 183296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.699050946596583e-05,
+      "loss": 4.4279,
+      "step": 183808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.698212351845531e-05,
+      "loss": 4.4509,
+      "step": 184320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.697373757094479e-05,
+      "loss": 4.4166,
+      "step": 184832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6965351623434276e-05,
+      "loss": 4.4239,
+      "step": 185344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6956965675923756e-05,
+      "loss": 4.4303,
+      "step": 185856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.694857972841323e-05,
+      "loss": 4.4423,
+      "step": 186368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.694019378090271e-05,
+      "loss": 4.4291,
+      "step": 186880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6931824212195925e-05,
+      "loss": 4.4205,
+      "step": 187392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.69234382646854e-05,
+      "loss": 4.4233,
+      "step": 187904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.691505231717488e-05,
+      "loss": 4.4245,
+      "step": 188416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.690666636966436e-05,
+      "loss": 4.4341,
+      "step": 188928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6898296800957574e-05,
+      "loss": 4.4288,
+      "step": 189440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.688991085344705e-05,
+      "loss": 4.4278,
+      "step": 189952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.688154128474026e-05,
+      "loss": 4.433,
+      "step": 190464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.687315533722974e-05,
+      "loss": 4.436,
+      "step": 190976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.686476938971923e-05,
+      "loss": 4.4169,
+      "step": 191488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.68563834422087e-05,
+      "loss": 4.4169,
+      "step": 192000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.684799749469818e-05,
+      "loss": 4.4259,
+      "step": 192512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.68396279259914e-05,
+      "loss": 4.4087,
+      "step": 193024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.683124197848087e-05,
+      "loss": 4.4251,
+      "step": 193536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.682285603097035e-05,
+      "loss": 4.4103,
+      "step": 194048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.681447008345983e-05,
+      "loss": 4.4326,
+      "step": 194560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.680608413594931e-05,
+      "loss": 4.4184,
+      "step": 195072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.679771456724252e-05,
+      "loss": 4.4049,
+      "step": 195584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.6789328619732e-05,
+      "loss": 4.4091,
+      "step": 196096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.678094267222148e-05,
+      "loss": 4.4171,
+      "step": 196608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.677255672471097e-05,
+      "loss": 4.4258,
+      "step": 197120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.676417077720045e-05,
+      "loss": 4.4099,
+      "step": 197632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.675578482968993e-05,
+      "loss": 4.4148,
+      "step": 198144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.674739888217941e-05,
+      "loss": 4.4055,
+      "step": 198656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.673901293466889e-05,
+      "loss": 4.4166,
+      "step": 199168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6730643365962096e-05,
+      "loss": 4.4055,
+      "step": 199680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6722257418451576e-05,
+      "loss": 4.4021,
+      "step": 200192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6713871470941056e-05,
+      "loss": 4.4138,
+      "step": 200704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6705485523430536e-05,
+      "loss": 4.4107,
+      "step": 201216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6697099575920016e-05,
+      "loss": 4.4113,
+      "step": 201728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6688730007213225e-05,
+      "loss": 4.3995,
+      "step": 202240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6680344059702705e-05,
+      "loss": 4.3994,
+      "step": 202752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6671958112192185e-05,
+      "loss": 4.4021,
+      "step": 203264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6663572164681665e-05,
+      "loss": 4.404,
+      "step": 203776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.665520259597488e-05,
+      "loss": 4.4035,
+      "step": 204288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.664681664846436e-05,
+      "loss": 4.394,
+      "step": 204800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.663843070095384e-05,
+      "loss": 4.3916,
+      "step": 205312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.663004475344332e-05,
+      "loss": 4.4033,
+      "step": 205824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.662167518473653e-05,
+      "loss": 4.3754,
+      "step": 206336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.661328923722601e-05,
+      "loss": 4.3937,
+      "step": 206848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.660491966851922e-05,
+      "loss": 4.3852,
+      "step": 207360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.659655009981243e-05,
+      "loss": 4.3997,
+      "step": 207872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.658816415230191e-05,
+      "loss": 4.3893,
+      "step": 208384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.657977820479139e-05,
+      "loss": 4.3975,
+      "step": 208896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.657139225728087e-05,
+      "loss": 4.3876,
+      "step": 209408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6563006309770355e-05,
+      "loss": 4.386,
+      "step": 209920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6554620362259835e-05,
+      "loss": 4.397,
+      "step": 210432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6546234414749315e-05,
+      "loss": 4.3852,
+      "step": 210944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6537848467238795e-05,
+      "loss": 4.3961,
+      "step": 211456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6529462519728275e-05,
+      "loss": 4.3927,
+      "step": 211968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6521092951021484e-05,
+      "loss": 4.3928,
+      "step": 212480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6512707003510964e-05,
+      "loss": 4.3836,
+      "step": 212992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6504321056000444e-05,
+      "loss": 4.3802,
+      "step": 213504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6495935108489924e-05,
+      "loss": 4.3954,
+      "step": 214016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.3765,
+      "step": 214528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.647917959227261e-05,
+      "loss": 4.3928,
+      "step": 215040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.647079364476209e-05,
+      "loss": 4.3781,
+      "step": 215552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.646240769725157e-05,
+      "loss": 4.3835,
+      "step": 216064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.645402174974105e-05,
+      "loss": 4.3806,
+      "step": 216576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.644563580223053e-05,
+      "loss": 4.3797,
+      "step": 217088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.643726623352375e-05,
+      "loss": 4.3599,
+      "step": 217600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.642889666481696e-05,
+      "loss": 4.3796,
+      "step": 218112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.642051071730644e-05,
+      "loss": 4.3646,
+      "step": 218624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.641212476979592e-05,
+      "loss": 4.3811,
+      "step": 219136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.64037388222854e-05,
+      "loss": 4.3947,
+      "step": 219648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.639535287477488e-05,
+      "loss": 4.3884,
+      "step": 220160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.638696692726436e-05,
+      "loss": 4.3585,
+      "step": 220672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.637858097975383e-05,
+      "loss": 4.383,
+      "step": 221184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.637019503224331e-05,
+      "loss": 4.3781,
+      "step": 221696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.636180908473279e-05,
+      "loss": 4.3758,
+      "step": 222208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6353439516026006e-05,
+      "loss": 4.3673,
+      "step": 222720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6345053568515486e-05,
+      "loss": 4.3771,
+      "step": 223232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6336667621004966e-05,
+      "loss": 4.3712,
+      "step": 223744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6328281673494446e-05,
+      "loss": 4.3686,
+      "step": 224256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6319895725983926e-05,
+      "loss": 4.3699,
+      "step": 224768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6311509778473406e-05,
+      "loss": 4.3739,
+      "step": 225280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6303140209766615e-05,
+      "loss": 4.3683,
+      "step": 225792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6294754262256095e-05,
+      "loss": 4.3655,
+      "step": 226304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6286368314745575e-05,
+      "loss": 4.3719,
+      "step": 226816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6277982367235055e-05,
+      "loss": 4.3701,
+      "step": 227328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6269596419724535e-05,
+      "loss": 4.3623,
+      "step": 227840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6261210472214015e-05,
+      "loss": 4.378,
+      "step": 228352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.6252824524703495e-05,
+      "loss": 4.363,
+      "step": 228864
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.3638200759887695,
+      "eval_runtime": 293.4397,
+      "eval_samples_per_second": 1300.407,
+      "eval_steps_per_second": 40.639,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6244438577192975e-05,
+      "loss": 4.3519,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623605262968246e-05,
+      "loss": 4.3561,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622766668217194e-05,
+      "loss": 4.3736,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621928073466142e-05,
+      "loss": 4.3614,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6210894787150895e-05,
+      "loss": 4.3746,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6202508839640374e-05,
+      "loss": 4.3481,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6194122892129854e-05,
+      "loss": 4.3681,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3374,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6177350997108814e-05,
+      "loss": 4.37,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.3507,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160579102087774e-05,
+      "loss": 4.3643,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6152193154577254e-05,
+      "loss": 4.3637,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614382358587046e-05,
+      "loss": 4.3506,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613543763835994e-05,
+      "loss": 4.3544,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612705169084942e-05,
+      "loss": 4.347,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611866574333891e-05,
+      "loss": 4.3402,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611027979582839e-05,
+      "loss": 4.3536,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.3501,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.3526,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3601,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3661,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3537,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6059980489569e-05,
+      "loss": 4.3455,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605159454205848e-05,
+      "loss": 4.3597,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604320859454796e-05,
+      "loss": 4.343,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603482264703743e-05,
+      "loss": 4.3574,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602643669952691e-05,
+      "loss": 4.3517,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601805075201639e-05,
+      "loss": 4.3427,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600966480450588e-05,
+      "loss": 4.339,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600129523579909e-05,
+      "loss": 4.3438,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599290928828857e-05,
+      "loss": 4.3411,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598452334077805e-05,
+      "loss": 4.3479,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3493,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.3457,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595936549824649e-05,
+      "loss": 4.3466,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595097955073597e-05,
+      "loss": 4.3438,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594259360322545e-05,
+      "loss": 4.328,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5934224034518656e-05,
+      "loss": 4.3347,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5925838087008136e-05,
+      "loss": 4.3273,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5917452139497616e-05,
+      "loss": 4.3334,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5909066191987096e-05,
+      "loss": 4.3347,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590071300208404e-05,
+      "loss": 4.3399,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589232705457352e-05,
+      "loss": 4.3338,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5883941107063e-05,
+      "loss": 4.3271,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587555515955248e-05,
+      "loss": 4.3409,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586716921204196e-05,
+      "loss": 4.3348,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3402,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3326,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584202774831413e-05,
+      "loss": 4.3182,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583364180080361e-05,
+      "loss": 4.3354,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582525585329309e-05,
+      "loss": 4.3239,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58168862845863e-05,
+      "loss": 4.3323,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5808500337075786e-05,
+      "loss": 4.3226,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5800114389565266e-05,
+      "loss": 4.3224,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5791728442054746e-05,
+      "loss": 4.3219,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3285,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.3237,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.324,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3247,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5749831462109604e-05,
+      "loss": 4.3159,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3134,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5733059567088564e-05,
+      "loss": 4.3365,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724673619578044e-05,
+      "loss": 4.3028,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716287672067524e-05,
+      "loss": 4.307,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5707901724557004e-05,
+      "loss": 4.3121,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699515777046484e-05,
+      "loss": 4.3291,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691129829535963e-05,
+      "loss": 4.3143,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5682743882025443e-05,
+      "loss": 4.3093,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674357934514923e-05,
+      "loss": 4.3052,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.566598836580814e-05,
+      "loss": 4.3115,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565760241829762e-05,
+      "loss": 4.3235,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.315,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564083052327657e-05,
+      "loss": 4.3144,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563244457576605e-05,
+      "loss": 4.3265,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3234,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561568905954874e-05,
+      "loss": 4.3063,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560730311203822e-05,
+      "loss": 4.3082,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559891716452771e-05,
+      "loss": 4.3193,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559053121701719e-05,
+      "loss": 4.2947,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3154,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.2996,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3249,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.3089,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548634237072046e-05,
+      "loss": 4.299,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540248289561526e-05,
+      "loss": 4.298,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.3115,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.3189,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.3057,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5506720878323175e-05,
+      "loss": 4.3038,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5498334930812655e-05,
+      "loss": 4.2977,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548994898330214e-05,
+      "loss": 4.3091,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548156303579162e-05,
+      "loss": 4.2989,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54731770882811e-05,
+      "loss": 4.2974,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546479114077058e-05,
+      "loss": 4.3083,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.308,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.304,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.2917,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.2993,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.2965,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541449183451119e-05,
+      "loss": 4.2998,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54061222658044e-05,
+      "loss": 4.2998,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539773631829388e-05,
+      "loss": 4.2901,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538935037078336e-05,
+      "loss": 4.291,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538096442327284e-05,
+      "loss": 4.2988,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372578475762326e-05,
+      "loss": 4.2754,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5364192528251806e-05,
+      "loss": 4.2947,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5355822959545015e-05,
+      "loss": 4.2803,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5347437012034495e-05,
+      "loss": 4.2987,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5339051064523975e-05,
+      "loss": 4.2887,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5330665117013455e-05,
+      "loss": 4.2948,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532227916950293e-05,
+      "loss": 4.2858,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531389322199241e-05,
+      "loss": 4.2887,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.530550727448189e-05,
+      "loss": 4.298,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529712132697137e-05,
+      "loss": 4.2817,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528875175826458e-05,
+      "loss": 4.2956,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5280365810754064e-05,
+      "loss": 4.2937,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527199624204728e-05,
+      "loss": 4.2911,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526361029453675e-05,
+      "loss": 4.2923,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.525522434702623e-05,
+      "loss": 4.278,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524685477831945e-05,
+      "loss": 4.2968,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523846883080893e-05,
+      "loss": 4.2835,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52300828832984e-05,
+      "loss": 4.2923,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522169693578788e-05,
+      "loss": 4.2811,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521331098827736e-05,
+      "loss": 4.2862,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520492504076684e-05,
+      "loss": 4.2841,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519653909325632e-05,
+      "loss": 4.28,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.2678,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517976719823528e-05,
+      "loss": 4.2788,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51713976295285e-05,
+      "loss": 4.2732,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516301168201798e-05,
+      "loss": 4.2853,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515462573450746e-05,
+      "loss": 4.2971,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514623978699694e-05,
+      "loss": 4.2927,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513785383948642e-05,
+      "loss": 4.2632,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51294678919759e-05,
+      "loss": 4.2918,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5121098323269106e-05,
+      "loss": 4.283,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5112712375758586e-05,
+      "loss": 4.2815,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5104326428248066e-05,
+      "loss": 4.2751,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5095940480737546e-05,
+      "loss": 4.2838,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5087554533227026e-05,
+      "loss": 4.274,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5079168585716506e-05,
+      "loss": 4.2822,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5070782638205986e-05,
+      "loss": 4.2744,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5062396690695466e-05,
+      "loss": 4.2791,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505402712198868e-05,
+      "loss": 4.2786,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504564117447816e-05,
+      "loss": 4.273,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503727160577137e-05,
+      "loss": 4.2772,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502888565826085e-05,
+      "loss": 4.2831,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502049971075033e-05,
+      "loss": 4.267,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501211376323981e-05,
+      "loss": 4.283,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500372781572929e-05,
+      "loss": 4.2777,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.280254364013672,
+      "eval_runtime": 297.5027,
+      "eval_samples_per_second": 1282.647,
+      "eval_steps_per_second": 40.084,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4995341868218764e-05,
+      "loss": 4.2672,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4986955920708244e-05,
+      "loss": 4.2652,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978569973197724e-05,
+      "loss": 4.2788,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497020040449093e-05,
+      "loss": 4.2729,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.496181445698042e-05,
+      "loss": 4.2857,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49534285094699e-05,
+      "loss": 4.2576,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494504256195938e-05,
+      "loss": 4.2749,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.493667299325259e-05,
+      "loss": 4.2559,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492828704574207e-05,
+      "loss": 4.2758,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491990109823155e-05,
+      "loss": 4.2644,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491151515072103e-05,
+      "loss": 4.2719,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490312920321051e-05,
+      "loss": 4.2776,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489475963450372e-05,
+      "loss": 4.2592,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.48863736869932e-05,
+      "loss": 4.2692,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487798773948268e-05,
+      "loss": 4.2595,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486960179197216e-05,
+      "loss": 4.2516,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486121584446164e-05,
+      "loss": 4.2633,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4852846275754853e-05,
+      "loss": 4.2644,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.484446032824433e-05,
+      "loss": 4.2645,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.483607438073381e-05,
+      "loss": 4.2751,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.482768843322329e-05,
+      "loss": 4.2809,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481930248571277e-05,
+      "loss": 4.2682,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481093291700598e-05,
+      "loss": 4.2624,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.480254696949546e-05,
+      "loss": 4.274,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.479416102198494e-05,
+      "loss": 4.2549,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.478577507447442e-05,
+      "loss": 4.2693,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.47773891269639e-05,
+      "loss": 4.2674,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476900317945338e-05,
+      "loss": 4.2542,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476061723194286e-05,
+      "loss": 4.2567,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475223128443234e-05,
+      "loss": 4.257,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474386171572556e-05,
+      "loss": 4.2581,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473547576821504e-05,
+      "loss": 4.263,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472708982070452e-05,
+      "loss": 4.2663,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471872025199773e-05,
+      "loss": 4.2612,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471033430448721e-05,
+      "loss": 4.2636,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470194835697669e-05,
+      "loss": 4.2612,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469356240946617e-05,
+      "loss": 4.2439,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468517646195565e-05,
+      "loss": 4.2542,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467679051444512e-05,
+      "loss": 4.2446,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.46684045669346e-05,
+      "loss": 4.2495,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466001861942408e-05,
+      "loss": 4.25,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4651649050717296e-05,
+      "loss": 4.2626,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4643263103206776e-05,
+      "loss": 4.251,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4634877155696256e-05,
+      "loss": 4.2442,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4626491208185736e-05,
+      "loss": 4.2608,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.461812163947895e-05,
+      "loss": 4.251,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460975207077216e-05,
+      "loss": 4.2565,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460136612326164e-05,
+      "loss": 4.2548,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.459298017575112e-05,
+      "loss": 4.2373,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4584594228240594e-05,
+      "loss": 4.2558,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576208280730074e-05,
+      "loss": 4.2419,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4567822333219554e-05,
+      "loss": 4.2545,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4559436385709033e-05,
+      "loss": 4.2431,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4551050438198513e-05,
+      "loss": 4.2426,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.454266449068799e-05,
+      "loss": 4.2447,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4534311300784945e-05,
+      "loss": 4.2479,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4525925353274425e-05,
+      "loss": 4.244,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45175394057639e-05,
+      "loss": 4.2443,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450915345825338e-05,
+      "loss": 4.2488,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450076751074286e-05,
+      "loss": 4.2369,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.449238156323234e-05,
+      "loss": 4.2332,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.448399561572182e-05,
+      "loss": 4.2633,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.447562604701503e-05,
+      "loss": 4.2228,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446724009950451e-05,
+      "loss": 4.2295,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445885415199399e-05,
+      "loss": 4.2359,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445046820448347e-05,
+      "loss": 4.2531,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.444208225697295e-05,
+      "loss": 4.2312,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.443371268826616e-05,
+      "loss": 4.2362,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.442532674075564e-05,
+      "loss": 4.2256,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.441694079324512e-05,
+      "loss": 4.2352,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44085548457346e-05,
+      "loss": 4.2438,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440016889822408e-05,
+      "loss": 4.2421,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.439178295071356e-05,
+      "loss": 4.2389,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.438341338200677e-05,
+      "loss": 4.2468,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.437502743449625e-05,
+      "loss": 4.2458,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.436664148698573e-05,
+      "loss": 4.2326,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435825553947521e-05,
+      "loss": 4.2317,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434988597076842e-05,
+      "loss": 4.2451,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43415000232579e-05,
+      "loss": 4.2188,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433311407574738e-05,
+      "loss": 4.2389,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432472812823687e-05,
+      "loss": 4.2256,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431634218072635e-05,
+      "loss": 4.2496,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.430795623321583e-05,
+      "loss": 4.2343,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429957028570531e-05,
+      "loss": 4.2271,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429118433819479e-05,
+      "loss": 4.2183,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4282814769487996e-05,
+      "loss": 4.2394,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4274428821977476e-05,
+      "loss": 4.2409,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4266042874466956e-05,
+      "loss": 4.2362,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.425765692695643e-05,
+      "loss": 4.2287,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4249287358249645e-05,
+      "loss": 4.2273,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4240901410739125e-05,
+      "loss": 4.2333,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4232531842032334e-05,
+      "loss": 4.2241,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422414589452182e-05,
+      "loss": 4.2246,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.42157599470113e-05,
+      "loss": 4.2372,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.420737399950078e-05,
+      "loss": 4.2329,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419898805199026e-05,
+      "loss": 4.2314,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4190602104479734e-05,
+      "loss": 4.2222,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4182216156969214e-05,
+      "loss": 4.2275,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4173830209458694e-05,
+      "loss": 4.221,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.41654606407519e-05,
+      "loss": 4.2291,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.415707469324138e-05,
+      "loss": 4.2265,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414868874573086e-05,
+      "loss": 4.2235,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414030279822034e-05,
+      "loss": 4.2216,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.413193322951356e-05,
+      "loss": 4.2208,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.412354728200304e-05,
+      "loss": 4.2159,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411516133449252e-05,
+      "loss": 4.2182,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4106775386982e-05,
+      "loss": 4.2105,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409838943947148e-05,
+      "loss": 4.2279,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409000349196096e-05,
+      "loss": 4.2207,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408163392325417e-05,
+      "loss": 4.2201,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407324797574365e-05,
+      "loss": 4.2145,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406486202823313e-05,
+      "loss": 4.224,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405647608072261e-05,
+      "loss": 4.2251,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404809013321209e-05,
+      "loss": 4.2128,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403970418570157e-05,
+      "loss": 4.2264,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403131823819105e-05,
+      "loss": 4.2263,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4022948669484257e-05,
+      "loss": 4.2218,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401456272197374e-05,
+      "loss": 4.2254,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400617677446322e-05,
+      "loss": 4.2096,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39977908269527e-05,
+      "loss": 4.224,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398942125824591e-05,
+      "loss": 4.2194,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398103531073539e-05,
+      "loss": 4.2258,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397264936322487e-05,
+      "loss": 4.2056,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396426341571435e-05,
+      "loss": 4.2226,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.395589384700756e-05,
+      "loss": 4.2175,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394750789949704e-05,
+      "loss": 4.2107,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393912195198652e-05,
+      "loss": 4.2028,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3930736004476e-05,
+      "loss": 4.2105,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.392236643576921e-05,
+      "loss": 4.2062,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39139804882587e-05,
+      "loss": 4.2198,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390559454074818e-05,
+      "loss": 4.2256,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389720859323766e-05,
+      "loss": 4.2282,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388882264572714e-05,
+      "loss": 4.199,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388043669821662e-05,
+      "loss": 4.2242,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2185,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2101,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.2071,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3846909286978266e-05,
+      "loss": 4.2206,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383852333946774e-05,
+      "loss": 4.2097,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3830153770760955e-05,
+      "loss": 4.2138,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3821767823250435e-05,
+      "loss": 4.2055,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3813381875739915e-05,
+      "loss": 4.212,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.213,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3796609980718875e-05,
+      "loss": 4.2095,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378824041201209e-05,
+      "loss": 4.2125,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.2178,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.2039,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3763082569480524e-05,
+      "loss": 4.2148,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.375471300077374e-05,
+      "loss": 4.2145,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.223569393157959,
+      "eval_runtime": 297.3907,
+      "eval_samples_per_second": 1283.13,
+      "eval_steps_per_second": 40.099,
+      "step": 381600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.374632705326321e-05,
+      "loss": 4.2015,
+      "step": 381952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.373794110575269e-05,
+      "loss": 4.1987,
+      "step": 382464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.372955515824217e-05,
+      "loss": 4.2152,
+      "step": 382976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.372118558953539e-05,
+      "loss": 4.209,
+      "step": 383488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.371279964202487e-05,
+      "loss": 4.2164,
+      "step": 384000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.370441369451435e-05,
+      "loss": 4.2009,
+      "step": 384512
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.369602774700383e-05,
+      "loss": 4.2084,
+      "step": 385024
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.368764179949331e-05,
+      "loss": 4.1936,
+      "step": 385536
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.367925585198279e-05,
+      "loss": 4.2075,
+      "step": 386048
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.367086990447227e-05,
+      "loss": 4.205,
+      "step": 386560
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.366248395696175e-05,
+      "loss": 4.2085,
+      "step": 387072
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.365411438825496e-05,
+      "loss": 4.211,
+      "step": 387584
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3645744819548166e-05,
+      "loss": 4.2004,
+      "step": 388096
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3637358872037646e-05,
+      "loss": 4.2017,
+      "step": 388608
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3628972924527126e-05,
+      "loss": 4.1981,
+      "step": 389120
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3620586977016606e-05,
+      "loss": 4.1891,
+      "step": 389632
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3612201029506086e-05,
+      "loss": 4.1968,
+      "step": 390144
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3603815081995566e-05,
+      "loss": 4.2024,
+      "step": 390656
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.359544551328878e-05,
+      "loss": 4.2019,
+      "step": 391168
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.358705956577826e-05,
+      "loss": 4.2133,
+      "step": 391680
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.357867361826774e-05,
+      "loss": 4.2175,
+      "step": 392192
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.357028767075722e-05,
+      "loss": 4.21,
+      "step": 392704
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.35619017232467e-05,
+      "loss": 4.2,
+      "step": 393216
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.355353215453991e-05,
+      "loss": 4.2104,
+      "step": 393728
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.354514620702939e-05,
+      "loss": 4.1982,
+      "step": 394240
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.353676025951887e-05,
+      "loss": 4.2066,
+      "step": 394752
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.352837431200835e-05,
+      "loss": 4.2047,
+      "step": 395264
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.351998836449783e-05,
+      "loss": 4.1902,
+      "step": 395776
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.351160241698731e-05,
+      "loss": 4.194,
+      "step": 396288
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.350321646947679e-05,
+      "loss": 4.2006,
+      "step": 396800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3494846900770006e-05,
+      "loss": 4.1986,
+      "step": 397312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3486460953259486e-05,
+      "loss": 4.2009,
+      "step": 397824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3478075005748966e-05,
+      "loss": 4.2065,
+      "step": 398336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3469689058238446e-05,
+      "loss": 4.1991,
+      "step": 398848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3461319489531655e-05,
+      "loss": 4.2022,
+      "step": 399360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3452933542021135e-05,
+      "loss": 4.2019,
+      "step": 399872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3444547594510615e-05,
+      "loss": 4.187,
+      "step": 400384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3436161647000095e-05,
+      "loss": 4.1877,
+      "step": 400896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3427775699489575e-05,
+      "loss": 4.1847,
+      "step": 401408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.341938975197905e-05,
+      "loss": 4.1897,
+      "step": 401920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.341100380446853e-05,
+      "loss": 4.1905,
+      "step": 402432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.340261785695801e-05,
+      "loss": 4.2032,
+      "step": 402944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.339423190944749e-05,
+      "loss": 4.191,
+      "step": 403456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3385862340740704e-05,
+      "loss": 4.1852,
+      "step": 403968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3377476393230184e-05,
+      "loss": 4.205,
+      "step": 404480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.33691068245234e-05,
+      "loss": 4.1877,
+      "step": 404992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.336072087701287e-05,
+      "loss": 4.2018,
+      "step": 405504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.335233492950235e-05,
+      "loss": 4.1937,
+      "step": 406016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.334394898199183e-05,
+      "loss": 4.1809,
+      "step": 406528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.333556303448131e-05,
+      "loss": 4.1963,
+      "step": 407040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.332717708697079e-05,
+      "loss": 4.186,
+      "step": 407552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.331879113946027e-05,
+      "loss": 4.1991,
+      "step": 408064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.331040519194975e-05,
+      "loss": 4.1806,
+      "step": 408576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.330203562324296e-05,
+      "loss": 4.1848,
+      "step": 409088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.329364967573244e-05,
+      "loss": 4.1831,
+      "step": 409600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.328528010702566e-05,
+      "loss": 4.1888,
+      "step": 410112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.327689415951514e-05,
+      "loss": 4.1895,
+      "step": 410624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.326850821200462e-05,
+      "loss": 4.1813,
+      "step": 411136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.32601222644941e-05,
+      "loss": 4.1932,
+      "step": 411648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.325173631698358e-05,
+      "loss": 4.1831,
+      "step": 412160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.324335036947306e-05,
+      "loss": 4.1724,
+      "step": 412672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.323496442196254e-05,
+      "loss": 4.2074,
+      "step": 413184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.322657847445202e-05,
+      "loss": 4.1651,
+      "step": 413696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3218225284548956e-05,
+      "loss": 4.1655,
+      "step": 414208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3209839337038436e-05,
+      "loss": 4.1798,
+      "step": 414720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3201453389527916e-05,
+      "loss": 4.1989,
+      "step": 415232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3193067442017396e-05,
+      "loss": 4.1718,
+      "step": 415744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.318469787331061e-05,
+      "loss": 4.1841,
+      "step": 416256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.317631192580009e-05,
+      "loss": 4.1687,
+      "step": 416768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.316792597828957e-05,
+      "loss": 4.178,
+      "step": 417280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.315954003077905e-05,
+      "loss": 4.1901,
+      "step": 417792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.315115408326853e-05,
+      "loss": 4.1869,
+      "step": 418304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.314276813575801e-05,
+      "loss": 4.1833,
+      "step": 418816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.313438218824749e-05,
+      "loss": 4.1868,
+      "step": 419328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.312599624073697e-05,
+      "loss": 4.1941,
+      "step": 419840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.311761029322645e-05,
+      "loss": 4.1736,
+      "step": 420352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.310922434571593e-05,
+      "loss": 4.1805,
+      "step": 420864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.310085477700914e-05,
+      "loss": 4.1867,
+      "step": 421376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.309246882949862e-05,
+      "loss": 4.1664,
+      "step": 421888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.30840828819881e-05,
+      "loss": 4.1804,
+      "step": 422400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.307569693447758e-05,
+      "loss": 4.176,
+      "step": 422912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.306731098696706e-05,
+      "loss": 4.1898,
+      "step": 423424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.305892503945654e-05,
+      "loss": 4.1822,
+      "step": 423936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.305053909194602e-05,
+      "loss": 4.1713,
+      "step": 424448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.30421531444355e-05,
+      "loss": 4.1638,
+      "step": 424960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.303378357572871e-05,
+      "loss": 4.1832,
+      "step": 425472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.302539762821819e-05,
+      "loss": 4.1837,
+      "step": 425984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.301701168070767e-05,
+      "loss": 4.177,
+      "step": 426496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3008642112000885e-05,
+      "loss": 4.1788,
+      "step": 427008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.300025616449036e-05,
+      "loss": 4.1742,
+      "step": 427520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.299187021697984e-05,
+      "loss": 4.1785,
+      "step": 428032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.298348426946932e-05,
+      "loss": 4.1701,
+      "step": 428544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.29750983219588e-05,
+      "loss": 4.1696,
+      "step": 429056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2966728753252014e-05,
+      "loss": 4.1847,
+      "step": 429568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2958342805741494e-05,
+      "loss": 4.1799,
+      "step": 430080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2949956858230974e-05,
+      "loss": 4.1755,
+      "step": 430592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2941570910720454e-05,
+      "loss": 4.167,
+      "step": 431104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2933184963209934e-05,
+      "loss": 4.1764,
+      "step": 431616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2924799015699413e-05,
+      "loss": 4.1647,
+      "step": 432128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2916413068188893e-05,
+      "loss": 4.1775,
+      "step": 432640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.29080434994821e-05,
+      "loss": 4.1742,
+      "step": 433152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.289965755197158e-05,
+      "loss": 4.1712,
+      "step": 433664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.289128798326479e-05,
+      "loss": 4.1688,
+      "step": 434176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.288290203575427e-05,
+      "loss": 4.1608,
+      "step": 434688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.287451608824375e-05,
+      "loss": 4.1682,
+      "step": 435200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.286613014073324e-05,
+      "loss": 4.1616,
+      "step": 435712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.285774419322272e-05,
+      "loss": 4.1626,
+      "step": 436224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.28493582457122e-05,
+      "loss": 4.1718,
+      "step": 436736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.284097229820168e-05,
+      "loss": 4.1706,
+      "step": 437248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.283258635069116e-05,
+      "loss": 4.1674,
+      "step": 437760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.282420040318064e-05,
+      "loss": 4.1594,
+      "step": 438272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.281581445567012e-05,
+      "loss": 4.1694,
+      "step": 438784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.28074285081596e-05,
+      "loss": 4.1731,
+      "step": 439296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.279904256064907e-05,
+      "loss": 4.1632,
+      "step": 439808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.279067299194229e-05,
+      "loss": 4.1761,
+      "step": 440320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.278228704443177e-05,
+      "loss": 4.1718,
+      "step": 440832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.277390109692124e-05,
+      "loss": 4.1724,
+      "step": 441344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.276551514941072e-05,
+      "loss": 4.1701,
+      "step": 441856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2757145580703936e-05,
+      "loss": 4.1592,
+      "step": 442368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2748759633193416e-05,
+      "loss": 4.1673,
+      "step": 442880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2740373685682896e-05,
+      "loss": 4.17,
+      "step": 443392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2731987738172376e-05,
+      "loss": 4.1747,
+      "step": 443904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2723601790661856e-05,
+      "loss": 4.1577,
+      "step": 444416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2715215843151336e-05,
+      "loss": 4.168,
+      "step": 444928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2706829895640816e-05,
+      "loss": 4.1688,
+      "step": 445440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2698443948130296e-05,
+      "loss": 4.1607,
+      "step": 445952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2690074379423505e-05,
+      "loss": 4.1487,
+      "step": 446464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2681688431912985e-05,
+      "loss": 4.1635,
+      "step": 446976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2673302484402465e-05,
+      "loss": 4.1532,
+      "step": 447488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2664916536891945e-05,
+      "loss": 4.1688,
+      "step": 448000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2656546968185154e-05,
+      "loss": 4.1741,
+      "step": 448512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.264816102067464e-05,
+      "loss": 4.1765,
+      "step": 449024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.263977507316412e-05,
+      "loss": 4.148,
+      "step": 449536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.263140550445733e-05,
+      "loss": 4.1728,
+      "step": 450048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.262301955694681e-05,
+      "loss": 4.1672,
+      "step": 450560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.261463360943629e-05,
+      "loss": 4.1596,
+      "step": 451072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.260624766192577e-05,
+      "loss": 4.1583,
+      "step": 451584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.259786171441525e-05,
+      "loss": 4.173,
+      "step": 452096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.258949214570846e-05,
+      "loss": 4.1554,
+      "step": 452608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.258110619819794e-05,
+      "loss": 4.167,
+      "step": 453120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.257273662949115e-05,
+      "loss": 4.1544,
+      "step": 453632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.256435068198063e-05,
+      "loss": 4.1651,
+      "step": 454144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.255596473447011e-05,
+      "loss": 4.1624,
+      "step": 454656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2547578786959594e-05,
+      "loss": 4.1597,
+      "step": 455168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2539192839449074e-05,
+      "loss": 4.1673,
+      "step": 455680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2530806891938554e-05,
+      "loss": 4.1611,
+      "step": 456192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2522420944428034e-05,
+      "loss": 4.1625,
+      "step": 456704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2514034996917514e-05,
+      "loss": 4.1584,
+      "step": 457216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2505649049406994e-05,
+      "loss": 4.1703,
+      "step": 457728
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.182224750518799,
+      "eval_runtime": 296.6792,
+      "eval_samples_per_second": 1286.208,
+      "eval_steps_per_second": 40.195,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497263101896474e-05,
+      "loss": 4.1577,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488877154385954e-05,
+      "loss": 4.1487,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480491206875434e-05,
+      "loss": 4.1613,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247210525936491e-05,
+      "loss": 4.1642,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246371931185439e-05,
+      "loss": 4.1669,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245533336434387e-05,
+      "loss": 4.1527,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244694741683335e-05,
+      "loss": 4.1579,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243856146932283e-05,
+      "loss": 4.1455,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243017552181231e-05,
+      "loss": 4.161,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2421789574301793e-05,
+      "loss": 4.1541,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241340362679127e-05,
+      "loss": 4.1639,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240501767928075e-05,
+      "loss": 4.16,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239664811057396e-05,
+      "loss": 4.1539,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238826216306344e-05,
+      "loss": 4.1576,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237987621555292e-05,
+      "loss": 4.1451,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23714902680424e-05,
+      "loss": 4.1408,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236310432053188e-05,
+      "loss": 4.1502,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235471837302136e-05,
+      "loss": 4.1585,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234634880431457e-05,
+      "loss": 4.1525,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233796285680405e-05,
+      "loss": 4.1701,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232957690929353e-05,
+      "loss": 4.1659,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232119096178301e-05,
+      "loss": 4.1642,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231282139307623e-05,
+      "loss": 4.1532,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230443544556571e-05,
+      "loss": 4.16,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229604949805519e-05,
+      "loss": 4.148,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228766355054467e-05,
+      "loss": 4.1636,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227927760303415e-05,
+      "loss": 4.1545,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2270908034327356e-05,
+      "loss": 4.144,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262522086816836e-05,
+      "loss": 4.1519,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254136139306316e-05,
+      "loss": 4.1513,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224575019179579e-05,
+      "loss": 4.1513,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223736424428527e-05,
+      "loss": 4.1568,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2228994675578485e-05,
+      "loss": 4.1562,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220608728067965e-05,
+      "loss": 4.16,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212222780557445e-05,
+      "loss": 4.1481,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203836833046925e-05,
+      "loss": 4.154,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195450885536405e-05,
+      "loss": 4.149,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187064938025885e-05,
+      "loss": 4.137,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178678990515365e-05,
+      "loss": 4.1392,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170293043004845e-05,
+      "loss": 4.1429,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2161907095494325e-05,
+      "loss": 4.1443,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2153521147983804e-05,
+      "loss": 4.158,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145135200473284e-05,
+      "loss": 4.1461,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2136749252962764e-05,
+      "loss": 4.1412,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128379684255973e-05,
+      "loss": 4.1607,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212001011554918e-05,
+      "loss": 4.1415,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211162416803866e-05,
+      "loss": 4.1542,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210323822052815e-05,
+      "loss": 4.1476,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209485227301763e-05,
+      "loss": 4.1283,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208646632550711e-05,
+      "loss": 4.1575,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207808037799659e-05,
+      "loss": 4.1387,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206969443048607e-05,
+      "loss": 4.1529,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206130848297555e-05,
+      "loss": 4.1392,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205295529307249e-05,
+      "loss": 4.1409,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204456934556197e-05,
+      "loss": 4.1331,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036199776855176e-05,
+      "loss": 4.1416,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027813829344656e-05,
+      "loss": 4.1498,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019427881834136e-05,
+      "loss": 4.133,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011041934323616e-05,
+      "loss": 4.1496,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20026559868131e-05,
+      "loss": 4.1409,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199427003930258e-05,
+      "loss": 4.1289,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198588409179206e-05,
+      "loss": 4.1569,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197749814428154e-05,
+      "loss": 4.1243,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196911219677102e-05,
+      "loss": 4.1224,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19607262492605e-05,
+      "loss": 4.1389,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195234030174998e-05,
+      "loss": 4.1505,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1943954354239456e-05,
+      "loss": 4.1277,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193558478553267e-05,
+      "loss": 4.1432,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192719883802215e-05,
+      "loss": 4.122,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918812890511625e-05,
+      "loss": 4.1381,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910426943001105e-05,
+      "loss": 4.1428,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190205737429432e-05,
+      "loss": 4.1406,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18936714267838e-05,
+      "loss": 4.1406,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188528547927328e-05,
+      "loss": 4.1394,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187689953176276e-05,
+      "loss": 4.151,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868529963055976e-05,
+      "loss": 4.1303,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860144015545456e-05,
+      "loss": 4.1384,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851774446838665e-05,
+      "loss": 4.1378,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843388499328145e-05,
+      "loss": 4.1251,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835002551817625e-05,
+      "loss": 4.1371,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826632983110834e-05,
+      "loss": 4.1334,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818247035600314e-05,
+      "loss": 4.1467,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809861088089794e-05,
+      "loss": 4.1443,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801475140579274e-05,
+      "loss": 4.1224,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793089193068754e-05,
+      "loss": 4.1235,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784703245558234e-05,
+      "loss": 4.1401,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776317298047714e-05,
+      "loss": 4.1445,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767931350537194e-05,
+      "loss": 4.1371,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17595617818304e-05,
+      "loss": 4.1301,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175117583431988e-05,
+      "loss": 4.1343,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174278988680936e-05,
+      "loss": 4.1332,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173440393929884e-05,
+      "loss": 4.1312,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172601799178832e-05,
+      "loss": 4.1253,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17176320442778e-05,
+      "loss": 4.139,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170926247557101e-05,
+      "loss": 4.1355,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170087652806049e-05,
+      "loss": 4.137,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169249058054998e-05,
+      "loss": 4.126,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168410463303946e-05,
+      "loss": 4.1313,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167571868552894e-05,
+      "loss": 4.1271,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166733273801842e-05,
+      "loss": 4.1348,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16589467905079e-05,
+      "loss": 4.1317,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165056084299738e-05,
+      "loss": 4.1255,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164219127429059e-05,
+      "loss": 4.1327,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163380532678007e-05,
+      "loss": 4.1211,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162541937926955e-05,
+      "loss": 4.1249,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161703343175903e-05,
+      "loss": 4.1206,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160864748424851e-05,
+      "loss": 4.1241,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160026153673799e-05,
+      "loss": 4.1236,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159187558922746e-05,
+      "loss": 4.1341,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158348964171694e-05,
+      "loss": 4.1192,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575120073010156e-05,
+      "loss": 4.123,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156673412549964e-05,
+      "loss": 4.1247,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558348177989116e-05,
+      "loss": 4.1301,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1549962230478596e-05,
+      "loss": 4.1245,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154159266177181e-05,
+      "loss": 4.1369,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533206714261285e-05,
+      "loss": 4.1295,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524820766750765e-05,
+      "loss": 4.1297,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151645119804398e-05,
+      "loss": 4.1294,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150806525053346e-05,
+      "loss": 4.1215,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1499679303022934e-05,
+      "loss": 4.125,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1491293355512414e-05,
+      "loss": 4.1305,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482907408001894e-05,
+      "loss": 4.1269,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147452146049138e-05,
+      "loss": 4.1201,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146613551298086e-05,
+      "loss": 4.1311,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145774956547034e-05,
+      "loss": 4.1258,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144936361795982e-05,
+      "loss": 4.1222,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14409776704493e-05,
+      "loss": 4.1086,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143260810174251e-05,
+      "loss": 4.1206,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142422215423199e-05,
+      "loss": 4.1132,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141583620672147e-05,
+      "loss": 4.1288,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140745025921095e-05,
+      "loss": 4.1346,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139908069050416e-05,
+      "loss": 4.1341,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139069474299364e-05,
+      "loss": 4.1082,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138230879548312e-05,
+      "loss": 4.1327,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13739228479726e-05,
+      "loss": 4.1264,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136553690046208e-05,
+      "loss": 4.1237,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357150952951565e-05,
+      "loss": 4.1172,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1348765005441045e-05,
+      "loss": 4.132,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340379057930525e-05,
+      "loss": 4.1146,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332009489223734e-05,
+      "loss": 4.1311,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323623541713214e-05,
+      "loss": 4.1123,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315237594202694e-05,
+      "loss": 4.1255,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306851646692174e-05,
+      "loss": 4.1217,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129849845678911e-05,
+      "loss": 4.1193,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129011250927859e-05,
+      "loss": 4.1245,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128172656176807e-05,
+      "loss": 4.1224,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127334061425755e-05,
+      "loss": 4.1219,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12649874243545e-05,
+      "loss": 4.1185,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125660147684398e-05,
+      "loss": 4.133,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.151233673095703,
+      "eval_runtime": 295.0048,
+      "eval_samples_per_second": 1293.508,
+      "eval_steps_per_second": 40.423,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124821552933346e-05,
+      "loss": 4.1152,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123982958182294e-05,
+      "loss": 4.1088,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123144363431242e-05,
+      "loss": 4.1262,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12230576868019e-05,
+      "loss": 4.1241,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121467173929138e-05,
+      "loss": 4.1268,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120628579178086e-05,
+      "loss": 4.1162,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119789984427034e-05,
+      "loss": 4.1205,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118951389675982e-05,
+      "loss": 4.1066,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11811279492493e-05,
+      "loss": 4.1211,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117274200173877e-05,
+      "loss": 4.1183,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116435605422826e-05,
+      "loss": 4.121,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115597010671774e-05,
+      "loss": 4.1215,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114761691681468e-05,
+      "loss": 4.1185,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113923096930416e-05,
+      "loss": 4.1137,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113084502179364e-05,
+      "loss": 4.1077,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112245907428312e-05,
+      "loss": 4.1015,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114073126772595e-05,
+      "loss": 4.1133,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105687179262075e-05,
+      "loss": 4.1179,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1097301231751555e-05,
+      "loss": 4.113,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1088915284241035e-05,
+      "loss": 4.1247,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1080529336730515e-05,
+      "loss": 4.1303,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072143389219995e-05,
+      "loss": 4.1285,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063757441709475e-05,
+      "loss": 4.1151,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1055371494198955e-05,
+      "loss": 4.1189,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1046985546688435e-05,
+      "loss": 4.1169,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103861597798165e-05,
+      "loss": 4.1211,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103023003047113e-05,
+      "loss": 4.1146,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102184408296061e-05,
+      "loss": 4.1124,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101347451425382e-05,
+      "loss": 4.1154,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10050885667433e-05,
+      "loss": 4.1121,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099670261923278e-05,
+      "loss": 4.1096,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098831667172226e-05,
+      "loss": 4.1183,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097993072421174e-05,
+      "loss": 4.1216,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097154477670122e-05,
+      "loss": 4.1201,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096317520799443e-05,
+      "loss": 4.1097,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095478926048391e-05,
+      "loss": 4.1204,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094640331297339e-05,
+      "loss": 4.1151,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938033744266604e-05,
+      "loss": 4.0976,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0929647796756084e-05,
+      "loss": 4.102,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921261849245564e-05,
+      "loss": 4.1059,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0912875901735044e-05,
+      "loss": 4.1114,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0904489954224524e-05,
+      "loss": 4.1159,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896104006714004e-05,
+      "loss": 4.1096,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0887718059203484e-05,
+      "loss": 4.1017,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087933211169296e-05,
+      "loss": 4.1254,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087094616418244e-05,
+      "loss": 4.1052,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086256021667192e-05,
+      "loss": 4.1186,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085419064796513e-05,
+      "loss": 4.1138,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084580470045461e-05,
+      "loss": 4.086,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083741875294409e-05,
+      "loss": 4.125,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082903280543357e-05,
+      "loss": 4.0994,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082064685792305e-05,
+      "loss": 4.1152,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081226091041253e-05,
+      "loss": 4.1055,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080387496290201e-05,
+      "loss": 4.1013,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079548901539149e-05,
+      "loss": 4.1017,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07871194466847e-05,
+      "loss": 4.1014,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077874987797791e-05,
+      "loss": 4.1191,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077036393046739e-05,
+      "loss": 4.0928,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076197798295687e-05,
+      "loss": 4.1125,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075360841425008e-05,
+      "loss": 4.1094,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0745222466739566e-05,
+      "loss": 4.0909,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0736836519229046e-05,
+      "loss": 4.1172,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0728450571718526e-05,
+      "loss": 4.0943,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0720064624208006e-05,
+      "loss": 4.0893,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0711678676697486e-05,
+      "loss": 4.1002,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0703292729186966e-05,
+      "loss": 4.1109,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0694906781676446e-05,
+      "loss": 4.0943,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0686520834165926e-05,
+      "loss": 4.1089,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678134886655406e-05,
+      "loss": 4.0867,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0669748939144886e-05,
+      "loss": 4.0992,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0661362991634366e-05,
+      "loss": 4.1085,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0652993422927575e-05,
+      "loss": 4.1021,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0644607475417055e-05,
+      "loss": 4.1086,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636221527906535e-05,
+      "loss": 4.1022,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627835580396015e-05,
+      "loss": 4.1127,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619449632885495e-05,
+      "loss": 4.0984,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061108006417871e-05,
+      "loss": 4.1043,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060269411666819e-05,
+      "loss": 4.0969,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059430816915767e-05,
+      "loss": 4.0954,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058595497925461e-05,
+      "loss": 4.1005,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057756903174409e-05,
+      "loss": 4.1001,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056918308423357e-05,
+      "loss": 4.1083,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056079713672305e-05,
+      "loss": 4.1124,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055241118921253e-05,
+      "loss": 4.0848,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054402524170201e-05,
+      "loss": 4.0888,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053563929419149e-05,
+      "loss": 4.1023,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052725334668097e-05,
+      "loss": 4.109,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051886739917045e-05,
+      "loss": 4.1014,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0510497830463664e-05,
+      "loss": 4.0961,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0502111882953144e-05,
+      "loss": 4.1042,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049372593544262e-05,
+      "loss": 4.0952,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04853399879321e-05,
+      "loss": 4.0982,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047695404042158e-05,
+      "loss": 4.0909,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046856809291106e-05,
+      "loss": 4.1029,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460198524204267e-05,
+      "loss": 4.1042,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0451812576693746e-05,
+      "loss": 4.1046,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443426629183226e-05,
+      "loss": 4.0925,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435040681672706e-05,
+      "loss": 4.0928,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0426654734162186e-05,
+      "loss": 4.0913,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418268786651666e-05,
+      "loss": 4.1014,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040988283914115e-05,
+      "loss": 4.0994,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040151327043436e-05,
+      "loss": 4.0901,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039312732292384e-05,
+      "loss": 4.0965,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038474137541332e-05,
+      "loss": 4.088,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03763554279028e-05,
+      "loss": 4.089,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036798585919601e-05,
+      "loss": 4.0848,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035959991168549e-05,
+      "loss": 4.0927,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035121396417497e-05,
+      "loss": 4.0912,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034282801666445e-05,
+      "loss": 4.0967,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033445844795766e-05,
+      "loss": 4.0906,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032607250044714e-05,
+      "loss": 4.0838,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031768655293662e-05,
+      "loss": 4.0935,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030930060542611e-05,
+      "loss": 4.0944,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0300931036719316e-05,
+      "loss": 4.0927,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0292545089208796e-05,
+      "loss": 4.101,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284159141698276e-05,
+      "loss": 4.0937,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0275773194187756e-05,
+      "loss": 4.1031,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267403625480965e-05,
+      "loss": 4.0936,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259017677970445e-05,
+      "loss": 4.0885,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0250631730459925e-05,
+      "loss": 4.0898,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242245782949405e-05,
+      "loss": 4.1031,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0233859835438885e-05,
+      "loss": 4.0915,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0225473887928365e-05,
+      "loss": 4.0871,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217104319221574e-05,
+      "loss": 4.0966,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020871837171106e-05,
+      "loss": 4.0956,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020033242420054e-05,
+      "loss": 4.0858,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019194647669002e-05,
+      "loss": 4.0794,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018357690798323e-05,
+      "loss": 4.0876,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017519096047271e-05,
+      "loss": 4.0816,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016680501296219e-05,
+      "loss": 4.0954,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015841906545167e-05,
+      "loss": 4.1015,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015003311794115e-05,
+      "loss": 4.1029,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014164717043063e-05,
+      "loss": 4.0751,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01332612229201e-05,
+      "loss": 4.1009,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012487527540958e-05,
+      "loss": 4.0927,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01165057067028e-05,
+      "loss": 4.0909,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010811975919228e-05,
+      "loss": 4.0849,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009973381168176e-05,
+      "loss": 4.0974,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009134786417124e-05,
+      "loss": 4.0854,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008296191666072e-05,
+      "loss": 4.1009,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007459234795393e-05,
+      "loss": 4.0793,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006620640044341e-05,
+      "loss": 4.0917,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005782045293289e-05,
+      "loss": 4.091,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004943450542237e-05,
+      "loss": 4.0891,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004104855791185e-05,
+      "loss": 4.0911,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0032678989205056e-05,
+      "loss": 4.0887,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024293041694536e-05,
+      "loss": 4.0922,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0015907094184016e-05,
+      "loss": 4.0885,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000753752547723e-05,
+      "loss": 4.0977,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.12639856338501,
+      "eval_runtime": 294.6039,
+      "eval_samples_per_second": 1295.268,
+      "eval_steps_per_second": 40.478,
+      "step": 610560
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.999915157796671e-05,
+      "loss": 4.0801,
+      "step": 610816
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.999076563045619e-05,
+      "loss": 4.0796,
+      "step": 611328
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.998237968294567e-05,
+      "loss": 4.0896,
+      "step": 611840
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.997399373543515e-05,
+      "loss": 4.0926,
+      "step": 612352
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.996560778792463e-05,
+      "loss": 4.0973,
+      "step": 612864
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.995722184041411e-05,
+      "loss": 4.0832,
+      "step": 613376
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.994883589290359e-05,
+      "loss": 4.0927,
+      "step": 613888
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.99404663241968e-05,
+      "loss": 4.0774,
+      "step": 614400
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.993208037668628e-05,
+      "loss": 4.0843,
+      "step": 614912
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.992369442917576e-05,
+      "loss": 4.0876,
+      "step": 615424
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.991530848166524e-05,
+      "loss": 4.0899,
+      "step": 615936
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.990693891295845e-05,
+      "loss": 4.0886,
+      "step": 616448
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9898569344251665e-05,
+      "loss": 4.0906,
+      "step": 616960
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9890183396741145e-05,
+      "loss": 4.0788,
+      "step": 617472
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9881797449230625e-05,
+      "loss": 4.0784,
+      "step": 617984
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9873411501720105e-05,
+      "loss": 4.0706,
+      "step": 618496
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9865025554209585e-05,
+      "loss": 4.0823,
+      "step": 619008
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9856655985502794e-05,
+      "loss": 4.0914,
+      "step": 619520
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9848270037992274e-05,
+      "loss": 4.0783,
+      "step": 620032
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9839884090481754e-05,
+      "loss": 4.0934,
+      "step": 620544
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9831498142971234e-05,
+      "loss": 4.0977,
+      "step": 621056
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9823112195460714e-05,
+      "loss": 4.0954,
+      "step": 621568
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9814726247950194e-05,
+      "loss": 4.0904,
+      "step": 622080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9806340300439674e-05,
+      "loss": 4.082,
+      "step": 622592
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.9797954352929154e-05,
+      "loss": 4.0836,
+      "step": 623104
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.97896011630261e-05,
+      "loss": 4.0896,
+      "step": 623616
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.978121521551558e-05,
+      "loss": 4.0898,
+      "step": 624128
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.977282926800506e-05,
+      "loss": 4.08,
+      "step": 624640
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.976444332049454e-05,
+      "loss": 4.0851,
+      "step": 625152
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.975605737298402e-05,
+      "loss": 4.0817,
+      "step": 625664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.97476714254735e-05,
+      "loss": 4.0773,
+      "step": 626176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.973928547796298e-05,
+      "loss": 4.0864,
+      "step": 626688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.973089953045246e-05,
+      "loss": 4.0876,
+      "step": 627200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.972252996174567e-05,
+      "loss": 4.0903,
+      "step": 627712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.971416039303888e-05,
+      "loss": 4.0801,
+      "step": 628224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.970579082433209e-05,
+      "loss": 4.088,
+      "step": 628736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.969740487682157e-05,
+      "loss": 4.0864,
+      "step": 629248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.968901892931105e-05,
+      "loss": 4.0638,
+      "step": 629760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.968063298180053e-05,
+      "loss": 4.0761,
+      "step": 630272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.967224703429001e-05,
+      "loss": 4.0704,
+      "step": 630784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.966386108677949e-05,
+      "loss": 4.0773,
+      "step": 631296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.965547513926897e-05,
+      "loss": 4.0906,
+      "step": 631808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.964708919175845e-05,
+      "loss": 4.0778,
+      "step": 632320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.963871962305166e-05,
+      "loss": 4.074,
+      "step": 632832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.963033367554114e-05,
+      "loss": 4.088,
+      "step": 633344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.962194772803062e-05,
+      "loss": 4.0767,
+      "step": 633856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.96135617805201e-05,
+      "loss": 4.0909,
+      "step": 634368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.960519221181331e-05,
+      "loss": 4.0831,
+      "step": 634880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.959680626430279e-05,
+      "loss": 4.057,
+      "step": 635392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.958842031679228e-05,
+      "loss": 4.0938,
+      "step": 635904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.958003436928176e-05,
+      "loss": 4.0709,
+      "step": 636416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.957164842177124e-05,
+      "loss": 4.0841,
+      "step": 636928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.956326247426071e-05,
+      "loss": 4.0752,
+      "step": 637440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.955487652675019e-05,
+      "loss": 4.0722,
+      "step": 637952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9546506958043406e-05,
+      "loss": 4.0704,
+      "step": 638464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.953812101053288e-05,
+      "loss": 4.0727,
+      "step": 638976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.952973506302236e-05,
+      "loss": 4.0916,
+      "step": 639488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.952134911551184e-05,
+      "loss": 4.062,
+      "step": 640000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.951296316800132e-05,
+      "loss": 4.0881,
+      "step": 640512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.95045772204908e-05,
+      "loss": 4.0772,
+      "step": 641024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.949619127298028e-05,
+      "loss": 4.0582,
+      "step": 641536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.948780532546976e-05,
+      "loss": 4.0887,
+      "step": 642048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9479435756762975e-05,
+      "loss": 4.0667,
+      "step": 642560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9471049809252455e-05,
+      "loss": 4.0557,
+      "step": 643072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9462663861741935e-05,
+      "loss": 4.0695,
+      "step": 643584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9454294293035144e-05,
+      "loss": 4.08,
+      "step": 644096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9445908345524624e-05,
+      "loss": 4.0673,
+      "step": 644608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9437522398014104e-05,
+      "loss": 4.0776,
+      "step": 645120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9429136450503584e-05,
+      "loss": 4.0612,
+      "step": 645632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9420750502993064e-05,
+      "loss": 4.0664,
+      "step": 646144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9412364555482544e-05,
+      "loss": 4.0786,
+      "step": 646656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.9403978607972024e-05,
+      "loss": 4.0759,
+      "step": 647168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.939560903926523e-05,
+      "loss": 4.0764,
+      "step": 647680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.938722309175471e-05,
+      "loss": 4.072,
+      "step": 648192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.93788371442442e-05,
+      "loss": 4.0875,
+      "step": 648704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.937045119673368e-05,
+      "loss": 4.0671,
+      "step": 649216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.936206524922316e-05,
+      "loss": 4.0771,
+      "step": 649728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.93537120593201e-05,
+      "loss": 4.0663,
+      "step": 650240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.934532611180958e-05,
+      "loss": 4.0659,
+      "step": 650752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.933694016429906e-05,
+      "loss": 4.0746,
+      "step": 651264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.932855421678854e-05,
+      "loss": 4.068,
+      "step": 651776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.932016826927802e-05,
+      "loss": 4.0776,
+      "step": 652288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.93117823217675e-05,
+      "loss": 4.0874,
+      "step": 652800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.930339637425698e-05,
+      "loss": 4.0556,
+      "step": 653312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.929501042674646e-05,
+      "loss": 4.0592,
+      "step": 653824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.928662447923594e-05,
+      "loss": 4.0751,
+      "step": 654336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.927823853172542e-05,
+      "loss": 4.0764,
+      "step": 654848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.92698525842149e-05,
+      "loss": 4.0744,
+      "step": 655360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.926148301550811e-05,
+      "loss": 4.0686,
+      "step": 655872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.925309706799759e-05,
+      "loss": 4.0733,
+      "step": 656384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9244711120487066e-05,
+      "loss": 4.0685,
+      "step": 656896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9236325172976546e-05,
+      "loss": 4.0655,
+      "step": 657408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9227939225466026e-05,
+      "loss": 4.0661,
+      "step": 657920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9219553277955506e-05,
+      "loss": 4.0756,
+      "step": 658432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9211183709248715e-05,
+      "loss": 4.0771,
+      "step": 658944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9202797761738195e-05,
+      "loss": 4.0749,
+      "step": 659456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9194411814227675e-05,
+      "loss": 4.0616,
+      "step": 659968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9186025866717155e-05,
+      "loss": 4.0666,
+      "step": 660480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9177639919206635e-05,
+      "loss": 4.0633,
+      "step": 660992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9169253971696115e-05,
+      "loss": 4.0712,
+      "step": 661504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.91608680241856e-05,
+      "loss": 4.0724,
+      "step": 662016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.915249845547881e-05,
+      "loss": 4.0596,
+      "step": 662528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.914411250796829e-05,
+      "loss": 4.0705,
+      "step": 663040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.913572656045777e-05,
+      "loss": 4.0547,
+      "step": 663552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.912734061294725e-05,
+      "loss": 4.062,
+      "step": 664064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.911897104424046e-05,
+      "loss": 4.0557,
+      "step": 664576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.911060147553367e-05,
+      "loss": 4.066,
+      "step": 665088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.910221552802315e-05,
+      "loss": 4.0643,
+      "step": 665600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.909382958051263e-05,
+      "loss": 4.0686,
+      "step": 666112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.908544363300211e-05,
+      "loss": 4.0593,
+      "step": 666624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.907705768549159e-05,
+      "loss": 4.0591,
+      "step": 667136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.906867173798107e-05,
+      "loss": 4.0607,
+      "step": 667648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9060285790470555e-05,
+      "loss": 4.0701,
+      "step": 668160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9051899842960035e-05,
+      "loss": 4.0636,
+      "step": 668672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9043513895449515e-05,
+      "loss": 4.0681,
+      "step": 669184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9035127947938995e-05,
+      "loss": 4.0659,
+      "step": 669696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9026742000428475e-05,
+      "loss": 4.0765,
+      "step": 670208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9018372431721684e-05,
+      "loss": 4.0665,
+      "step": 670720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9009986484211164e-05,
+      "loss": 4.0597,
+      "step": 671232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.9001600536700644e-05,
+      "loss": 4.0657,
+      "step": 671744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.8993214589190124e-05,
+      "loss": 4.0724,
+      "step": 672256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.8984828641679604e-05,
+      "loss": 4.0638,
+      "step": 672768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.897644269416908e-05,
+      "loss": 4.0598,
+      "step": 673280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.896805674665856e-05,
+      "loss": 4.066,
+      "step": 673792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.895967079914804e-05,
+      "loss": 4.0687,
+      "step": 674304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.895130123044125e-05,
+      "loss": 4.0586,
+      "step": 674816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.894291528293073e-05,
+      "loss": 4.052,
+      "step": 675328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.893454571422395e-05,
+      "loss": 4.0596,
+      "step": 675840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.892615976671343e-05,
+      "loss": 4.0523,
+      "step": 676352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.89177738192029e-05,
+      "loss": 4.0695,
+      "step": 676864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.890938787169238e-05,
+      "loss": 4.0699,
+      "step": 677376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.890100192418186e-05,
+      "loss": 4.0732,
+      "step": 677888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.889261597667134e-05,
+      "loss": 4.0493,
+      "step": 678400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.888423002916082e-05,
+      "loss": 4.0732,
+      "step": 678912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.88758440816503e-05,
+      "loss": 4.065,
+      "step": 679424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.886747451294351e-05,
+      "loss": 4.0636,
+      "step": 679936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.885908856543299e-05,
+      "loss": 4.0602,
+      "step": 680448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.885070261792248e-05,
+      "loss": 4.061,
+      "step": 680960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.884231667041196e-05,
+      "loss": 4.0603,
+      "step": 681472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.883393072290144e-05,
+      "loss": 4.0736,
+      "step": 681984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.882554477539092e-05,
+      "loss": 4.0511,
+      "step": 682496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.88171588278804e-05,
+      "loss": 4.0656,
+      "step": 683008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.880877288036988e-05,
+      "loss": 4.0657,
+      "step": 683520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.8800403311663086e-05,
+      "loss": 4.0639,
+      "step": 684032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.8792033742956296e-05,
+      "loss": 4.0589,
+      "step": 684544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.8783647795445775e-05,
+      "loss": 4.0627,
+      "step": 685056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.8775261847935255e-05,
+      "loss": 4.065,
+      "step": 685568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.8766875900424735e-05,
+      "loss": 4.0613,
+      "step": 686080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.8758489952914215e-05,
+      "loss": 4.0708,
+      "step": 686592
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.106845378875732,
+      "eval_runtime": 292.4243,
+      "eval_samples_per_second": 1304.922,
+      "eval_steps_per_second": 40.78,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8750104005403695e-05,
+      "loss": 4.0586,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8741718057893175e-05,
+      "loss": 4.0483,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8733332110382655e-05,
+      "loss": 4.0622,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872494616287214e-05,
+      "loss": 4.066,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8716560215361615e-05,
+      "loss": 4.0687,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8708174267851095e-05,
+      "loss": 4.0573,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8699788320340575e-05,
+      "loss": 4.0649,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8691402372830055e-05,
+      "loss": 4.0492,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8683016425319535e-05,
+      "loss": 4.058,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8674630477809015e-05,
+      "loss": 4.0619,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8666244530298495e-05,
+      "loss": 4.0642,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8657858582787975e-05,
+      "loss": 4.064,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8649489014081184e-05,
+      "loss": 4.0689,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8641103066570664e-05,
+      "loss": 4.0496,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8632717119060144e-05,
+      "loss": 4.0512,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8624331171549624e-05,
+      "loss": 4.0392,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861594522403911e-05,
+      "loss": 4.0595,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.860755927652859e-05,
+      "loss": 4.0654,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85991897078218e-05,
+      "loss": 4.0503,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859080376031128e-05,
+      "loss": 4.0647,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858241781280076e-05,
+      "loss": 4.0754,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857403186529024e-05,
+      "loss": 4.0669,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856564591777972e-05,
+      "loss": 4.062,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85572599702692e-05,
+      "loss": 4.0583,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854889040156241e-05,
+      "loss": 4.0616,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854050445405189e-05,
+      "loss": 4.0589,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853211850654137e-05,
+      "loss": 4.0632,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852373255903085e-05,
+      "loss": 4.0609,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851534661152033e-05,
+      "loss": 4.0543,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850696066400981e-05,
+      "loss": 4.0503,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849857471649929e-05,
+      "loss": 4.0516,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849018876898877e-05,
+      "loss": 4.0616,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.848180282147825e-05,
+      "loss": 4.0635,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.847341687396773e-05,
+      "loss": 4.0636,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846503092645721e-05,
+      "loss": 4.055,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845666135775042e-05,
+      "loss": 4.0575,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84482754102399e-05,
+      "loss": 4.0607,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843988946272938e-05,
+      "loss": 4.0397,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843150351521886e-05,
+      "loss": 4.048,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842311756770834e-05,
+      "loss": 4.0487,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841473162019782e-05,
+      "loss": 4.0488,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84063456726873e-05,
+      "loss": 4.0641,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839797610398051e-05,
+      "loss": 4.0537,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838959015646999e-05,
+      "loss": 4.0484,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838120420895947e-05,
+      "loss": 4.0634,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837281826144895e-05,
+      "loss": 4.048,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836443231393843e-05,
+      "loss": 4.06,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835604636642791e-05,
+      "loss": 4.0652,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834767679772112e-05,
+      "loss": 4.0308,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83392908502106e-05,
+      "loss": 4.0628,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833090490270008e-05,
+      "loss": 4.0493,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832251895518956e-05,
+      "loss": 4.0564,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831413300767904e-05,
+      "loss": 4.0497,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830576343897225e-05,
+      "loss": 4.044,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829737749146173e-05,
+      "loss": 4.0476,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8289007922754946e-05,
+      "loss": 4.0444,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8280621975244426e-05,
+      "loss": 4.0683,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8272236027733906e-05,
+      "loss": 4.0361,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8263850080223386e-05,
+      "loss": 4.0627,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8255496890320324e-05,
+      "loss": 4.0545,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8247110942809804e-05,
+      "loss": 4.0357,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8238724995299284e-05,
+      "loss": 4.0632,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8230339047788764e-05,
+      "loss": 4.0407,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8221953100278244e-05,
+      "loss": 4.0324,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8213567152767724e-05,
+      "loss": 4.0412,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8205181205257204e-05,
+      "loss": 4.0586,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8196795257746684e-05,
+      "loss": 4.0384,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8188409310236164e-05,
+      "loss": 4.0554,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8180023362725644e-05,
+      "loss": 4.0363,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171637415215124e-05,
+      "loss": 4.039,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816326784650834e-05,
+      "loss": 4.05,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.815488189899781e-05,
+      "loss": 4.0521,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814649595148729e-05,
+      "loss": 4.0538,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813811000397677e-05,
+      "loss": 4.0495,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812972405646625e-05,
+      "loss": 4.0571,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812133810895573e-05,
+      "loss": 4.0466,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811295216144521e-05,
+      "loss": 4.0507,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810458259273842e-05,
+      "loss": 4.037,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80961966452279e-05,
+      "loss": 4.0434,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808781069771739e-05,
+      "loss": 4.0485,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807942475020687e-05,
+      "loss": 4.0411,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807103880269635e-05,
+      "loss": 4.0541,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806265285518583e-05,
+      "loss": 4.0603,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805428328647904e-05,
+      "loss": 4.0396,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804589733896852e-05,
+      "loss": 4.0312,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8037511391458e-05,
+      "loss": 4.0482,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802912544394748e-05,
+      "loss": 4.0506,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802073949643696e-05,
+      "loss": 4.0502,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8012369927730166e-05,
+      "loss": 4.042,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8003983980219646e-05,
+      "loss": 4.0503,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7995598032709126e-05,
+      "loss": 4.0417,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7987212085198606e-05,
+      "loss": 4.0434,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7978826137688086e-05,
+      "loss": 4.0412,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79704565689813e-05,
+      "loss": 4.0484,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796207062147078e-05,
+      "loss": 4.0502,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795368467396026e-05,
+      "loss": 4.0518,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794529872644974e-05,
+      "loss": 4.0352,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793691277893922e-05,
+      "loss": 4.045,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79285268314287e-05,
+      "loss": 4.0367,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920140883918175e-05,
+      "loss": 4.0472,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911754936407655e-05,
+      "loss": 4.0465,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7903368988897135e-05,
+      "loss": 4.0395,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789499942019035e-05,
+      "loss": 4.046,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7886613472679824e-05,
+      "loss": 4.0307,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787824390397304e-05,
+      "loss": 4.0373,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786985795646253e-05,
+      "loss": 4.0317,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7861488387755736e-05,
+      "loss": 4.041,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7853102440245216e-05,
+      "loss": 4.0414,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7844716492734696e-05,
+      "loss": 4.0413,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7836330545224176e-05,
+      "loss": 4.0342,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782794459771365e-05,
+      "loss": 4.0345,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781955865020313e-05,
+      "loss": 4.0376,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781117270269261e-05,
+      "loss": 4.0462,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780278675518209e-05,
+      "loss": 4.0413,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779440080767157e-05,
+      "loss": 4.0421,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778601486016105e-05,
+      "loss": 4.0426,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777762891265053e-05,
+      "loss": 4.0494,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776924296514001e-05,
+      "loss": 4.0415,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7760873396433224e-05,
+      "loss": 4.0363,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7752487448922704e-05,
+      "loss": 4.0415,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7744117880215913e-05,
+      "loss": 4.0461,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7735731932705393e-05,
+      "loss": 4.0423,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772734598519487e-05,
+      "loss": 4.0371,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771896003768435e-05,
+      "loss": 4.0374,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771057409017383e-05,
+      "loss": 4.0464,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770218814266331e-05,
+      "loss": 4.0329,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769380219515279e-05,
+      "loss": 4.0278,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7685432626446e-05,
+      "loss": 4.0335,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767704667893548e-05,
+      "loss": 4.0252,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766866073142496e-05,
+      "loss": 4.0504,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766027478391444e-05,
+      "loss": 4.04,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765188883640393e-05,
+      "loss": 4.0534,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764351926769714e-05,
+      "loss": 4.028,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763513332018662e-05,
+      "loss": 4.0452,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762676375147983e-05,
+      "loss": 4.041,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761837780396931e-05,
+      "loss": 4.0413,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760999185645879e-05,
+      "loss": 4.0388,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760160590894827e-05,
+      "loss": 4.0375,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759321996143775e-05,
+      "loss": 4.0327,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758483401392723e-05,
+      "loss": 4.0533,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757644806641671e-05,
+      "loss": 4.0266,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756806211890618e-05,
+      "loss": 4.0427,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755967617139567e-05,
+      "loss": 4.0445,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755129022388515e-05,
+      "loss": 4.0379,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7542904276374627e-05,
+      "loss": 4.0347,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7534518328864107e-05,
+      "loss": 4.0392,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7526148760157316e-05,
+      "loss": 4.0432,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7517762812646796e-05,
+      "loss": 4.0411,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7509376865136276e-05,
+      "loss": 4.0454,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.090456962585449,
+      "eval_runtime": 295.0121,
+      "eval_samples_per_second": 1293.476,
+      "eval_steps_per_second": 40.422,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7500990917625755e-05,
+      "loss": 4.0269,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7492621348918965e-05,
+      "loss": 4.0271,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7484235401408445e-05,
+      "loss": 4.04,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7475849453897924e-05,
+      "loss": 4.0427,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7467463506387404e-05,
+      "loss": 4.0469,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7459077558876884e-05,
+      "loss": 4.0337,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7450691611366364e-05,
+      "loss": 4.0407,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744232204265958e-05,
+      "loss": 4.0307,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743393609514906e-05,
+      "loss": 4.0349,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742555014763854e-05,
+      "loss": 4.0377,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.741716420012802e-05,
+      "loss": 4.0368,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74087782526175e-05,
+      "loss": 4.0433,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740040868391071e-05,
+      "loss": 4.0433,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739202273640019e-05,
+      "loss": 4.0274,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738363678888967e-05,
+      "loss": 4.0289,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737525084137915e-05,
+      "loss": 4.0166,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736686489386863e-05,
+      "loss": 4.0401,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735847894635811e-05,
+      "loss": 4.0371,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735009299884759e-05,
+      "loss": 4.027,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734170705133707e-05,
+      "loss": 4.0378,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733332110382655e-05,
+      "loss": 4.0559,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732493515631603e-05,
+      "loss": 4.044,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7316565587609245e-05,
+      "loss": 4.0399,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7308179640098725e-05,
+      "loss": 4.0351,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.72997936925882e-05,
+      "loss": 4.041,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729140774507768e-05,
+      "loss": 4.0354,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7283038176370894e-05,
+      "loss": 4.0356,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7274652228860374e-05,
+      "loss": 4.0385,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726626628134985e-05,
+      "loss": 4.034,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725788033383933e-05,
+      "loss": 4.0309,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724949438632881e-05,
+      "loss": 4.031,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724112481762202e-05,
+      "loss": 4.0378,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.72327388701115e-05,
+      "loss": 4.0365,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722435292260098e-05,
+      "loss": 4.0419,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721596697509046e-05,
+      "loss": 4.0342,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720759740638367e-05,
+      "loss": 4.035,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719921145887315e-05,
+      "loss": 4.038,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719084189016637e-05,
+      "loss": 4.0209,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718245594265585e-05,
+      "loss": 4.0256,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717406999514532e-05,
+      "loss": 4.0266,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71656840476348e-05,
+      "loss": 4.0252,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7157314478928016e-05,
+      "loss": 4.0386,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7148928531417496e-05,
+      "loss": 4.0324,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7140542583906976e-05,
+      "loss": 4.0322,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7132156636396456e-05,
+      "loss": 4.0336,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712378706768967e-05,
+      "loss": 4.0285,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7115401120179145e-05,
+      "loss": 4.0392,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7107015172668625e-05,
+      "loss": 4.0425,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7098629225158105e-05,
+      "loss": 4.0069,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7090243277647585e-05,
+      "loss": 4.0409,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7081857330137065e-05,
+      "loss": 4.0249,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7073471382626545e-05,
+      "loss": 4.0365,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7065101813919754e-05,
+      "loss": 4.0294,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7056715866409234e-05,
+      "loss": 4.0187,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7048329918898714e-05,
+      "loss": 4.0289,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7039943971388194e-05,
+      "loss": 4.019,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7031558023877674e-05,
+      "loss": 4.0428,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702317207636716e-05,
+      "loss": 4.0181,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701478612885664e-05,
+      "loss": 4.037,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700640018134612e-05,
+      "loss": 4.032,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69980142338356e-05,
+      "loss": 4.0175,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698962828632508e-05,
+      "loss": 4.038,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698124233881456e-05,
+      "loss": 4.0166,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6972856391304034e-05,
+      "loss": 4.0131,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696448682259725e-05,
+      "loss": 4.0181,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695610087508673e-05,
+      "loss": 4.037,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69477149275762e-05,
+      "loss": 4.0181,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693932898006568e-05,
+      "loss": 4.0359,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693094303255516e-05,
+      "loss": 4.0105,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692257346384838e-05,
+      "loss": 4.0217,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691418751633786e-05,
+      "loss": 4.029,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690580156882734e-05,
+      "loss": 4.0298,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689741562131682e-05,
+      "loss": 4.0316,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6889046052610034e-05,
+      "loss": 4.0302,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688067648390324e-05,
+      "loss": 4.0351,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687229053639272e-05,
+      "loss": 4.0226,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68639045888822e-05,
+      "loss": 4.0292,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685553502017541e-05,
+      "loss": 4.0204,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.684714907266489e-05,
+      "loss": 4.0221,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683876312515437e-05,
+      "loss": 4.0247,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683037717764385e-05,
+      "loss": 4.0169,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682199123013333e-05,
+      "loss": 4.0313,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681360528262281e-05,
+      "loss": 4.0384,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680521933511229e-05,
+      "loss": 4.0244,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679683338760177e-05,
+      "loss": 4.006,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678844744009125e-05,
+      "loss": 4.0278,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678007787138446e-05,
+      "loss": 4.0285,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677169192387394e-05,
+      "loss": 4.033,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676330597636342e-05,
+      "loss": 4.0154,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675493640765663e-05,
+      "loss": 4.0309,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674655046014611e-05,
+      "loss": 4.0222,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673816451263559e-05,
+      "loss": 4.02,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672977856512507e-05,
+      "loss": 4.0195,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672139261761455e-05,
+      "loss": 4.0296,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6713006670104037e-05,
+      "loss": 4.0274,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6704620722593516e-05,
+      "loss": 4.0317,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6696234775082996e-05,
+      "loss": 4.0161,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6687848827572476e-05,
+      "loss": 4.023,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6679479258865685e-05,
+      "loss": 4.0153,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6671093311355165e-05,
+      "loss": 4.0243,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6662707363844645e-05,
+      "loss": 4.0239,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6654321416334125e-05,
+      "loss": 4.0205,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6645935468823605e-05,
+      "loss": 4.0226,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6637565900116814e-05,
+      "loss": 4.0087,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6629179952606294e-05,
+      "loss": 4.0212,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6620810383899503e-05,
+      "loss": 4.0082,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6612424436388983e-05,
+      "loss": 4.0208,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660403848887847e-05,
+      "loss": 4.0161,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.659565254136795e-05,
+      "loss": 4.0251,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658726659385743e-05,
+      "loss": 4.0133,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657888064634691e-05,
+      "loss": 4.015,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657049469883639e-05,
+      "loss": 4.0145,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.656210875132587e-05,
+      "loss": 4.0275,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.655372280381534e-05,
+      "loss": 4.0221,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.654533685630482e-05,
+      "loss": 4.0221,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65369509087943e-05,
+      "loss": 4.0234,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652858134008751e-05,
+      "loss": 4.0293,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652019539257699e-05,
+      "loss": 4.0207,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651180944506647e-05,
+      "loss": 4.0177,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650342349755595e-05,
+      "loss": 4.0191,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649503755004544e-05,
+      "loss": 4.0271,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648665160253492e-05,
+      "loss": 4.0203,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64782656550244e-05,
+      "loss": 4.0201,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646987970751388e-05,
+      "loss": 4.0182,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646149376000336e-05,
+      "loss": 4.0264,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645312419129657e-05,
+      "loss": 4.0113,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644473824378605e-05,
+      "loss": 4.0128,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643636867507926e-05,
+      "loss": 4.0078,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642798272756874e-05,
+      "loss": 4.0094,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6419596780058217e-05,
+      "loss": 4.0252,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6411210832547697e-05,
+      "loss": 4.0228,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6402824885037176e-05,
+      "loss": 4.034,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6394438937526656e-05,
+      "loss": 4.0069,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638606936881987e-05,
+      "loss": 4.0257,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637768342130935e-05,
+      "loss": 4.0189,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636929747379883e-05,
+      "loss": 4.0225,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636091152628831e-05,
+      "loss": 4.0161,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635252557877779e-05,
+      "loss": 4.0163,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634413963126727e-05,
+      "loss": 4.0161,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.633577006256048e-05,
+      "loss": 4.0322,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632738411504996e-05,
+      "loss": 4.0053,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631899816753944e-05,
+      "loss": 4.0235,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631061222002892e-05,
+      "loss": 4.0245,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63022262725184e-05,
+      "loss": 4.0163,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6293840325007874e-05,
+      "loss": 4.0153,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628545437749736e-05,
+      "loss": 4.0235,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627706842998684e-05,
+      "loss": 4.0194,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626869886128006e-05,
+      "loss": 4.0234,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6260329292573266e-05,
+      "loss": 4.0224,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.077542781829834,
+      "eval_runtime": 307.2271,
+      "eval_samples_per_second": 1242.049,
+      "eval_steps_per_second": 38.815,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6251943345062746e-05,
+      "loss": 3.9998,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6243557397552226e-05,
+      "loss": 4.0087,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.62351714500417e-05,
+      "loss": 4.0181,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622678550253118e-05,
+      "loss": 4.0221,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621839955502066e-05,
+      "loss": 4.0275,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621001360751014e-05,
+      "loss": 4.0172,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620162765999962e-05,
+      "loss": 4.0228,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.61932417124891e-05,
+      "loss": 4.0146,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618485576497858e-05,
+      "loss": 4.0116,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.617646981746806e-05,
+      "loss": 4.013,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.616808386995754e-05,
+      "loss": 4.0227,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6159697922447025e-05,
+      "loss": 4.0183,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6151328353740234e-05,
+      "loss": 4.0265,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6142942406229714e-05,
+      "loss": 4.0065,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6134556458719194e-05,
+      "loss": 4.009,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6126170511208674e-05,
+      "loss": 3.9969,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6117784563698154e-05,
+      "loss": 4.0198,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6109398616187634e-05,
+      "loss": 4.0176,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6101012668677114e-05,
+      "loss": 4.005,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609262672116659e-05,
+      "loss": 4.0211,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608424077365607e-05,
+      "loss": 4.0379,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607585482614555e-05,
+      "loss": 4.0224,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606748525743876e-05,
+      "loss": 4.0196,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605909930992824e-05,
+      "loss": 4.0166,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605071336241772e-05,
+      "loss": 4.021,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.60423274149072e-05,
+      "loss": 4.0183,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.603395784620042e-05,
+      "loss": 4.0152,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602557189868989e-05,
+      "loss": 4.0172,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601718595117937e-05,
+      "loss": 4.0153,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.600880000366885e-05,
+      "loss": 4.0114,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600041405615833e-05,
+      "loss": 4.0154,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.599202810864781e-05,
+      "loss": 4.0135,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598365853994102e-05,
+      "loss": 4.0178,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59752725924305e-05,
+      "loss": 4.0257,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596688664491998e-05,
+      "loss": 4.0146,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59585170762132e-05,
+      "loss": 4.0131,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595013112870268e-05,
+      "loss": 4.0209,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594174518119216e-05,
+      "loss": 4.001,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5933359233681637e-05,
+      "loss": 4.0094,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5924973286171117e-05,
+      "loss": 4.0052,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5916587338660597e-05,
+      "loss": 4.0092,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5908201391150076e-05,
+      "loss": 4.0122,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5899815443639556e-05,
+      "loss": 4.0132,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5891429496129036e-05,
+      "loss": 4.0154,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5883043548618516e-05,
+      "loss": 4.0141,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5874657601107996e-05,
+      "loss": 4.0151,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5866271653597476e-05,
+      "loss": 4.0151,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5857902084890685e-05,
+      "loss": 4.0243,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5849516137380165e-05,
+      "loss": 3.9911,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5841130189869645e-05,
+      "loss": 4.0235,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583274424235913e-05,
+      "loss": 4.007,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5824358294848605e-05,
+      "loss": 4.0153,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581598872614182e-05,
+      "loss": 4.0101,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58076027786313e-05,
+      "loss": 4.0054,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5799216831120774e-05,
+      "loss": 4.0035,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579084726241399e-05,
+      "loss": 3.9998,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578246131490347e-05,
+      "loss": 4.0288,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.577407536739295e-05,
+      "loss": 3.9982,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576568941988242e-05,
+      "loss": 4.0195,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57573034723719e-05,
+      "loss": 4.0071,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574893390366512e-05,
+      "loss": 4.001,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57405479561546e-05,
+      "loss": 4.0174,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573216200864408e-05,
+      "loss": 4.0027,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572377606113356e-05,
+      "loss": 3.9891,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571539011362304e-05,
+      "loss": 4.0012,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.570700416611252e-05,
+      "loss": 4.0188,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5698618218602e-05,
+      "loss": 3.9993,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569023227109148e-05,
+      "loss": 4.0174,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.568184632358096e-05,
+      "loss": 3.9903,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567346037607044e-05,
+      "loss": 4.0017,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566510718616738e-05,
+      "loss": 4.0098,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565672123865686e-05,
+      "loss": 4.0128,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564835166995007e-05,
+      "loss": 4.0112,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563996572243955e-05,
+      "loss": 4.0126,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563157977492903e-05,
+      "loss": 4.0183,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562319382741851e-05,
+      "loss": 4.0046,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.561480787990799e-05,
+      "loss": 4.0068,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560642193239747e-05,
+      "loss": 4.0018,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559803598488695e-05,
+      "loss": 4.0084,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558965003737643e-05,
+      "loss": 4.0036,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558126408986591e-05,
+      "loss": 4.0015,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557289452115912e-05,
+      "loss": 4.0138,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55645085736486e-05,
+      "loss": 4.0184,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555612262613808e-05,
+      "loss": 4.0064,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554773667862756e-05,
+      "loss": 3.9862,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553935073111704e-05,
+      "loss": 4.0089,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553098116241026e-05,
+      "loss": 4.0087,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552259521489974e-05,
+      "loss": 4.0172,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.551420926738922e-05,
+      "loss": 3.9982,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5505839698682426e-05,
+      "loss": 4.0105,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5497453751171906e-05,
+      "loss": 4.0103,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5489067803661386e-05,
+      "loss": 3.9981,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5480681856150866e-05,
+      "loss": 4.0001,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5472295908640346e-05,
+      "loss": 4.0135,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5463909961129826e-05,
+      "loss": 4.0066,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5455524013619306e-05,
+      "loss": 4.0139,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5447138066108786e-05,
+      "loss": 3.9999,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543875211859826e-05,
+      "loss": 4.0039,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5430382549891475e-05,
+      "loss": 3.9983,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5421996602380955e-05,
+      "loss": 4.0097,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.541361065487044e-05,
+      "loss": 4.0098,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5405224707359915e-05,
+      "loss": 3.9999,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5396838759849395e-05,
+      "loss": 4.0048,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5388452812338875e-05,
+      "loss": 3.9935,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5380083243632084e-05,
+      "loss": 4.0015,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5371697296121564e-05,
+      "loss": 3.9876,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5363311348611044e-05,
+      "loss": 4.0036,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5354925401100524e-05,
+      "loss": 3.9979,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5346539453590004e-05,
+      "loss": 4.0061,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533816988488321e-05,
+      "loss": 3.9996,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532978393737269e-05,
+      "loss": 4.0046,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532139798986218e-05,
+      "loss": 3.9924,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531301204235166e-05,
+      "loss": 4.0093,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530462609484114e-05,
+      "loss": 4.0025,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529624014733062e-05,
+      "loss": 4.0029,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52878541998201e-05,
+      "loss": 4.0027,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527946825230958e-05,
+      "loss": 4.0158,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527109868360279e-05,
+      "loss": 4.0004,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.526271273609227e-05,
+      "loss": 3.9995,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525432678858175e-05,
+      "loss": 4.0033,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524594084107123e-05,
+      "loss": 4.0079,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523757127236444e-05,
+      "loss": 4.0002,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522918532485392e-05,
+      "loss": 4.0049,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52207993773434e-05,
+      "loss": 3.999,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521241342983288e-05,
+      "loss": 4.005,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.520404386112609e-05,
+      "loss": 3.991,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.519565791361557e-05,
+      "loss": 3.9965,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518727196610505e-05,
+      "loss": 3.9898,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517888601859453e-05,
+      "loss": 3.9972,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517050007108401e-05,
+      "loss": 4.0053,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516211412357349e-05,
+      "loss": 4.0056,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515372817606297e-05,
+      "loss": 4.0133,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5145342228552446e-05,
+      "loss": 3.9938,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5136956281041926e-05,
+      "loss": 4.0041,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5128570333531406e-05,
+      "loss": 4.0057,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5120184386020886e-05,
+      "loss": 4.0043,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51118148173141e-05,
+      "loss": 3.9999,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510342886980358e-05,
+      "loss": 3.9946,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509504292229306e-05,
+      "loss": 4.0011,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508665697478254e-05,
+      "loss": 4.0098,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507827102727202e-05,
+      "loss": 3.9945,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506990145856523e-05,
+      "loss": 4.005,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506151551105471e-05,
+      "loss": 4.0058,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505312956354419e-05,
+      "loss": 4.0007,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50447599948374e-05,
+      "loss": 3.9968,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503637404732688e-05,
+      "loss": 4.0041,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502798809981636e-05,
+      "loss": 3.9997,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501960215230584e-05,
+      "loss": 4.0033,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501121620479532e-05,
+      "loss": 4.01,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.06660795211792,
+      "eval_runtime": 319.6273,
+      "eval_samples_per_second": 1193.862,
+      "eval_steps_per_second": 37.309,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5002846636088535e-05,
+      "loss": 3.9945,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4994460688578015e-05,
+      "loss": 3.9874,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4986074741067495e-05,
+      "loss": 4.0005,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4977705172360704e-05,
+      "loss": 4.0053,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4969319224850184e-05,
+      "loss": 4.0057,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4960933277339664e-05,
+      "loss": 4.0008,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4952547329829144e-05,
+      "loss": 4.0028,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4944161382318624e-05,
+      "loss": 3.9976,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4935775434808104e-05,
+      "loss": 3.9932,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4927389487297584e-05,
+      "loss": 3.9968,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4919003539787064e-05,
+      "loss": 4.0055,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4910617592276544e-05,
+      "loss": 3.9995,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.490224802356975e-05,
+      "loss": 4.0107,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.489386207605923e-05,
+      "loss": 3.9869,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488547612854872e-05,
+      "loss": 3.9933,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48770901810382e-05,
+      "loss": 3.9804,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486870423352768e-05,
+      "loss": 4.0022,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486031828601716e-05,
+      "loss": 3.9996,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.485193233850663e-05,
+      "loss": 3.9945,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.484356276979985e-05,
+      "loss": 3.9973,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.483517682228933e-05,
+      "loss": 4.0187,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482679087477881e-05,
+      "loss": 4.0057,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481840492726828e-05,
+      "loss": 4.002,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481001897975776e-05,
+      "loss": 3.9975,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.480164941105098e-05,
+      "loss": 4.0096,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.479326346354046e-05,
+      "loss": 3.9998,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.478487751602994e-05,
+      "loss": 3.9941,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.477649156851942e-05,
+      "loss": 4.0039,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.47681056210089e-05,
+      "loss": 3.9916,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4759736052302106e-05,
+      "loss": 3.998,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4751350104791586e-05,
+      "loss": 3.9989,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4742964157281066e-05,
+      "loss": 3.9946,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4734578209770546e-05,
+      "loss": 4.0007,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4726192262260026e-05,
+      "loss": 4.0106,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4717806314749506e-05,
+      "loss": 3.9954,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4709436746042715e-05,
+      "loss": 4.0002,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4701050798532195e-05,
+      "loss": 4.0045,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.469268122982541e-05,
+      "loss": 3.984,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468429528231489e-05,
+      "loss": 3.9898,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467590933480437e-05,
+      "loss": 3.9916,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466752338729385e-05,
+      "loss": 3.9891,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465913743978333e-05,
+      "loss": 3.9937,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465075149227281e-05,
+      "loss": 3.9975,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464236554476229e-05,
+      "loss": 3.9971,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46339959760555e-05,
+      "loss": 3.997,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462561002854498e-05,
+      "loss": 3.9969,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461722408103446e-05,
+      "loss": 4.0001,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460883813352394e-05,
+      "loss": 4.0046,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460045218601342e-05,
+      "loss": 3.9803,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45920662385029e-05,
+      "loss": 4.0026,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.458368029099238e-05,
+      "loss": 3.9888,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.457529434348186e-05,
+      "loss": 3.9963,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.456690839597134e-05,
+      "loss": 3.9958,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4558538827264555e-05,
+      "loss": 3.9885,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4550152879754035e-05,
+      "loss": 3.9877,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4541783311047244e-05,
+      "loss": 3.981,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4533397363536724e-05,
+      "loss": 4.0137,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4525011416026204e-05,
+      "loss": 3.9819,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4516625468515684e-05,
+      "loss": 3.9995,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4508239521005164e-05,
+      "loss": 3.991,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4499853573494644e-05,
+      "loss": 3.9855,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449146762598412e-05,
+      "loss": 3.9971,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44830816784736e-05,
+      "loss": 3.9885,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.447469573096308e-05,
+      "loss": 3.9773,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.446630978345256e-05,
+      "loss": 3.9775,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445794021474577e-05,
+      "loss": 4.0081,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444955426723525e-05,
+      "loss": 3.9853,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444116831972473e-05,
+      "loss": 3.9947,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443278237221421e-05,
+      "loss": 3.9778,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442441280350742e-05,
+      "loss": 3.9874,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44160268559969e-05,
+      "loss": 3.9884,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440764090848638e-05,
+      "loss": 3.9998,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439925496097586e-05,
+      "loss": 3.9946,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439086901346534e-05,
+      "loss": 3.9957,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.438248306595482e-05,
+      "loss": 4.0006,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.43740971184443e-05,
+      "loss": 3.9878,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.436571117093378e-05,
+      "loss": 3.9911,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435732522342326e-05,
+      "loss": 3.9839,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434897203352021e-05,
+      "loss": 3.9919,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434058608600969e-05,
+      "loss": 3.9897,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.433220013849917e-05,
+      "loss": 3.9822,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.432381419098865e-05,
+      "loss": 4.0002,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4315428243478127e-05,
+      "loss": 4.0019,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4307042295967607e-05,
+      "loss": 3.9914,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4298656348457086e-05,
+      "loss": 3.9664,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4290270400946566e-05,
+      "loss": 3.991,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4281900832239776e-05,
+      "loss": 3.9963,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4273514884729256e-05,
+      "loss": 3.9987,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4265128937218735e-05,
+      "loss": 3.9885,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.425675936851195e-05,
+      "loss": 3.9912,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424837342100143e-05,
+      "loss": 3.9925,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.423998747349091e-05,
+      "loss": 3.9833,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.423160152598039e-05,
+      "loss": 3.9849,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422321557846987e-05,
+      "loss": 3.9959,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.421482963095935e-05,
+      "loss": 3.9876,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.420646006225256e-05,
+      "loss": 4.0001,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.419807411474204e-05,
+      "loss": 3.9878,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418968816723152e-05,
+      "loss": 3.9849,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418131859852473e-05,
+      "loss": 3.9839,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.417293265101421e-05,
+      "loss": 3.9923,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.416454670350369e-05,
+      "loss": 3.9947,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.415616075599317e-05,
+      "loss": 3.9826,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414777480848265e-05,
+      "loss": 3.9882,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413938886097213e-05,
+      "loss": 3.9772,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413100291346161e-05,
+      "loss": 3.9902,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.412261696595109e-05,
+      "loss": 3.9679,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.411423101844057e-05,
+      "loss": 3.9845,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.410586144973378e-05,
+      "loss": 3.9869,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4097491881026994e-05,
+      "loss": 3.9903,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4089105933516474e-05,
+      "loss": 3.9774,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408071998600595e-05,
+      "loss": 3.99,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.407233403849543e-05,
+      "loss": 3.9769,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.406394809098491e-05,
+      "loss": 3.9926,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.405556214347439e-05,
+      "loss": 3.9853,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40471925747676e-05,
+      "loss": 3.9855,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403880662725708e-05,
+      "loss": 3.9855,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403042067974656e-05,
+      "loss": 4.0001,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.402205111103978e-05,
+      "loss": 3.9854,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.401366516352925e-05,
+      "loss": 3.9826,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.400527921601873e-05,
+      "loss": 3.989,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.399689326850821e-05,
+      "loss": 3.9872,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398850732099769e-05,
+      "loss": 3.9854,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398012137348717e-05,
+      "loss": 3.9883,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.397173542597665e-05,
+      "loss": 3.9822,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396334947846613e-05,
+      "loss": 3.9902,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.395496353095561e-05,
+      "loss": 3.975,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394657758344509e-05,
+      "loss": 3.9845,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393819163593457e-05,
+      "loss": 3.977,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392980568842405e-05,
+      "loss": 3.9842,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392141974091354e-05,
+      "loss": 3.9836,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391303379340302e-05,
+      "loss": 3.9913,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390466422469623e-05,
+      "loss": 3.9976,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389627827718571e-05,
+      "loss": 3.9799,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388789232967519e-05,
+      "loss": 3.9875,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387950638216467e-05,
+      "loss": 3.9869,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387112043465414e-05,
+      "loss": 3.9872,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3862750865947356e-05,
+      "loss": 3.9865,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3854364918436836e-05,
+      "loss": 3.9768,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.384597897092631e-05,
+      "loss": 3.989,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.383759302341579e-05,
+      "loss": 3.9941,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3829207075905276e-05,
+      "loss": 3.9767,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3820821128394756e-05,
+      "loss": 3.9876,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3812435180884236e-05,
+      "loss": 3.9925,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3804049233373716e-05,
+      "loss": 3.9801,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3795679664666925e-05,
+      "loss": 3.9822,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3787293717156405e-05,
+      "loss": 3.9913,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3778907769645885e-05,
+      "loss": 3.983,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3770521822135365e-05,
+      "loss": 3.9849,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3762152253428574e-05,
+      "loss": 3.996,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.056432247161865,
+      "eval_runtime": 295.56,
+      "eval_samples_per_second": 1291.078,
+      "eval_steps_per_second": 40.347,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3753766305918054e-05,
+      "loss": 3.988,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3745380358407534e-05,
+      "loss": 3.9687,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3736994410897014e-05,
+      "loss": 3.9833,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3728608463386493e-05,
+      "loss": 3.9939,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3720222515875973e-05,
+      "loss": 3.9888,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371183656836546e-05,
+      "loss": 3.9871,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370345062085494e-05,
+      "loss": 3.9837,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369506467334442e-05,
+      "loss": 3.9841,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36866787258339e-05,
+      "loss": 3.9772,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367829277832338e-05,
+      "loss": 3.9827,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366990683081285e-05,
+      "loss": 3.9862,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366152088330233e-05,
+      "loss": 3.9849,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365315131459555e-05,
+      "loss": 3.9958,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.364476536708503e-05,
+      "loss": 3.9712,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36363794195745e-05,
+      "loss": 3.9768,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362799347206398e-05,
+      "loss": 3.9643,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361960752455346e-05,
+      "loss": 3.9871,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361122157704294e-05,
+      "loss": 3.9841,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.360283562953243e-05,
+      "loss": 3.9801,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.359444968202191e-05,
+      "loss": 3.981,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.358606373451139e-05,
+      "loss": 4.0018,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357767778700087e-05,
+      "loss": 3.9906,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356930821829408e-05,
+      "loss": 3.9849,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356092227078356e-05,
+      "loss": 3.9815,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355253632327304e-05,
+      "loss": 3.9965,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354415037576252e-05,
+      "loss": 3.9807,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3535764428252e-05,
+      "loss": 3.9812,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352737848074148e-05,
+      "loss": 3.9886,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351899253323096e-05,
+      "loss": 3.9766,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351060658572044e-05,
+      "loss": 3.9813,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350222063820991e-05,
+      "loss": 3.9828,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3493851069503126e-05,
+      "loss": 3.9764,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3485465121992606e-05,
+      "loss": 3.9867,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347707917448209e-05,
+      "loss": 3.9925,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3468693226971566e-05,
+      "loss": 3.9804,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346032365826478e-05,
+      "loss": 3.9844,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345193771075426e-05,
+      "loss": 3.9936,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344355176324374e-05,
+      "loss": 3.9623,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3435165815733215e-05,
+      "loss": 3.9778,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342679624702643e-05,
+      "loss": 3.9779,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341841029951591e-05,
+      "loss": 3.9691,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3410024352005384e-05,
+      "loss": 3.9802,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3401638404494864e-05,
+      "loss": 3.9833,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.339326883578808e-05,
+      "loss": 3.9805,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338488288827756e-05,
+      "loss": 3.9814,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337649694076704e-05,
+      "loss": 3.9776,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3368127372060256e-05,
+      "loss": 3.9793,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3359741424549736e-05,
+      "loss": 3.9917,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3351371855842945e-05,
+      "loss": 3.9674,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3342985908332425e-05,
+      "loss": 3.9865,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3334599960821905e-05,
+      "loss": 3.9776,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3326214013311385e-05,
+      "loss": 3.9791,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331782806580086e-05,
+      "loss": 3.9807,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330944211829034e-05,
+      "loss": 3.9667,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330105617077982e-05,
+      "loss": 3.9794,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3292686602073034e-05,
+      "loss": 3.9644,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3284300654562514e-05,
+      "loss": 3.9964,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3275914707051994e-05,
+      "loss": 3.9649,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3267528759541474e-05,
+      "loss": 3.9865,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3259142812030954e-05,
+      "loss": 3.9736,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3250756864520434e-05,
+      "loss": 3.9702,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3242370917009914e-05,
+      "loss": 3.9784,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3233984969499393e-05,
+      "loss": 3.9755,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3225599021988873e-05,
+      "loss": 3.9654,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321721307447835e-05,
+      "loss": 3.9589,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320882712696783e-05,
+      "loss": 3.9891,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320045755826104e-05,
+      "loss": 3.9753,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.319207161075052e-05,
+      "loss": 3.9775,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318368566324e-05,
+      "loss": 3.9627,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317529971572948e-05,
+      "loss": 3.9696,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31669301470227e-05,
+      "loss": 3.9774,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315854419951218e-05,
+      "loss": 3.9842,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315015825200166e-05,
+      "loss": 3.9789,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314177230449114e-05,
+      "loss": 3.9828,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.313338635698062e-05,
+      "loss": 3.9836,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.312501678827383e-05,
+      "loss": 3.9728,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.311663084076331e-05,
+      "loss": 3.9741,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310824489325279e-05,
+      "loss": 3.9721,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309985894574227e-05,
+      "loss": 3.9753,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309147299823175e-05,
+      "loss": 3.9764,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3083103429524956e-05,
+      "loss": 3.9671,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3074717482014436e-05,
+      "loss": 3.9821,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3066331534503916e-05,
+      "loss": 3.9862,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30579455869934e-05,
+      "loss": 3.9741,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3049559639482876e-05,
+      "loss": 3.9563,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3041173691972356e-05,
+      "loss": 3.9754,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3032787744461836e-05,
+      "loss": 3.9798,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3024401796951316e-05,
+      "loss": 3.9807,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3016032228244525e-05,
+      "loss": 3.973,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300766265953774e-05,
+      "loss": 3.9744,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299927671202722e-05,
+      "loss": 3.9794,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2990890764516694e-05,
+      "loss": 3.9654,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2982504817006174e-05,
+      "loss": 3.9714,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2974118869495654e-05,
+      "loss": 3.9771,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296573292198514e-05,
+      "loss": 3.9734,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295736335327835e-05,
+      "loss": 3.9858,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2948993784571565e-05,
+      "loss": 3.9762,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2940607837061045e-05,
+      "loss": 3.9656,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293222188955052e-05,
+      "loss": 3.9676,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292383594204e-05,
+      "loss": 3.9768,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291544999452948e-05,
+      "loss": 3.9813,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.290706404701896e-05,
+      "loss": 3.9685,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289867809950844e-05,
+      "loss": 3.9699,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289029215199792e-05,
+      "loss": 3.962,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28819062044874e-05,
+      "loss": 3.9789,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287352025697688e-05,
+      "loss": 3.9534,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286513430946636e-05,
+      "loss": 3.9682,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285674836195584e-05,
+      "loss": 3.9729,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2848378793249054e-05,
+      "loss": 3.9766,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284000922454226e-05,
+      "loss": 3.9632,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283162327703174e-05,
+      "loss": 3.9731,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282323732952122e-05,
+      "loss": 3.9645,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28148513820107e-05,
+      "loss": 3.9764,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280646543450018e-05,
+      "loss": 3.9739,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279807948698966e-05,
+      "loss": 3.9677,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278969353947914e-05,
+      "loss": 3.973,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278130759196862e-05,
+      "loss": 3.9863,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.277293802326183e-05,
+      "loss": 3.9706,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276455207575131e-05,
+      "loss": 3.9723,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275616612824079e-05,
+      "loss": 3.972,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274779655953401e-05,
+      "loss": 3.9721,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273941061202349e-05,
+      "loss": 3.9721,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273102466451297e-05,
+      "loss": 3.9737,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.272263871700245e-05,
+      "loss": 3.9716,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.271425276949193e-05,
+      "loss": 3.9748,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270586682198141e-05,
+      "loss": 3.9631,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269748087447088e-05,
+      "loss": 3.9658,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268909492696036e-05,
+      "loss": 3.963,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2680725358253576e-05,
+      "loss": 3.9706,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2672339410743056e-05,
+      "loss": 3.9618,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266395346323253e-05,
+      "loss": 3.9836,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2655567515722016e-05,
+      "loss": 3.9836,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264719794701523e-05,
+      "loss": 3.9622,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263881199950471e-05,
+      "loss": 3.975,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2630426051994185e-05,
+      "loss": 3.973,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26220564832874e-05,
+      "loss": 3.9768,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261367053577688e-05,
+      "loss": 3.9657,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2605284588266354e-05,
+      "loss": 3.9664,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2596898640755834e-05,
+      "loss": 3.9764,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2588512693245314e-05,
+      "loss": 3.9793,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2580126745734794e-05,
+      "loss": 3.9627,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2571757177028e-05,
+      "loss": 3.9735,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.256337122951748e-05,
+      "loss": 3.9817,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255498528200697e-05,
+      "loss": 3.9676,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254659933449645e-05,
+      "loss": 3.9663,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253821338698593e-05,
+      "loss": 3.9767,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252982743947541e-05,
+      "loss": 3.9679,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252144149196489e-05,
+      "loss": 3.9744,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.251305554445437e-05,
+      "loss": 3.9765,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.049461364746094,
+      "eval_runtime": 352.8796,
+      "eval_samples_per_second": 1081.363,
+      "eval_steps_per_second": 33.793,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.250466959694385e-05,
+      "loss": 3.9697,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249628364943333e-05,
+      "loss": 3.955,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248791408072654e-05,
+      "loss": 3.9687,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247952813321602e-05,
+      "loss": 3.9828,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24711421857055e-05,
+      "loss": 3.9755,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246275623819498e-05,
+      "loss": 3.9735,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.245437029068446e-05,
+      "loss": 3.9686,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.244600072197767e-05,
+      "loss": 3.972,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243761477446715e-05,
+      "loss": 3.9633,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2429228826956634e-05,
+      "loss": 3.9735,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2420842879446114e-05,
+      "loss": 3.9681,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2412456931935594e-05,
+      "loss": 3.9758,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2404087363228803e-05,
+      "loss": 3.9806,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239570141571828e-05,
+      "loss": 3.9587,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238731546820776e-05,
+      "loss": 3.9638,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237892952069724e-05,
+      "loss": 3.95,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237055995199045e-05,
+      "loss": 3.971,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236217400447993e-05,
+      "loss": 3.973,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235378805696941e-05,
+      "loss": 3.966,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.234540210945889e-05,
+      "loss": 3.965,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233701616194837e-05,
+      "loss": 3.9923,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232863021443785e-05,
+      "loss": 3.9808,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232024426692733e-05,
+      "loss": 3.9707,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.231185831941681e-05,
+      "loss": 3.9708,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.230347237190629e-05,
+      "loss": 3.9855,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.229511918200324e-05,
+      "loss": 3.9693,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228673323449272e-05,
+      "loss": 3.9648,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227834728698219e-05,
+      "loss": 3.9796,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226996133947167e-05,
+      "loss": 3.9622,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226157539196115e-05,
+      "loss": 3.966,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.225318944445063e-05,
+      "loss": 3.9666,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.224480349694011e-05,
+      "loss": 3.9629,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2236433928233326e-05,
+      "loss": 3.9744,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2228047980722806e-05,
+      "loss": 3.9801,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2219662033212286e-05,
+      "loss": 3.9643,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2211292464505495e-05,
+      "loss": 3.9767,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2202906516994975e-05,
+      "loss": 3.9749,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2194520569484455e-05,
+      "loss": 3.9511,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2186134621973935e-05,
+      "loss": 3.9615,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2177765053267144e-05,
+      "loss": 3.9677,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2169379105756624e-05,
+      "loss": 3.9552,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2160993158246104e-05,
+      "loss": 3.9716,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2152607210735584e-05,
+      "loss": 3.9713,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2144221263225064e-05,
+      "loss": 3.9665,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2135835315714544e-05,
+      "loss": 3.969,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2127449368204024e-05,
+      "loss": 3.9693,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211906342069351e-05,
+      "loss": 3.9631,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211067747318299e-05,
+      "loss": 3.9805,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210229152567247e-05,
+      "loss": 3.9529,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209390557816195e-05,
+      "loss": 3.9714,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.208551963065143e-05,
+      "loss": 3.971,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20771336831409e-05,
+      "loss": 3.9646,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206876411443412e-05,
+      "loss": 3.968,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206039454572733e-05,
+      "loss": 3.9538,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205202497702054e-05,
+      "loss": 3.9667,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.204363902951002e-05,
+      "loss": 3.9489,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20352530819995e-05,
+      "loss": 3.9847,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202686713448898e-05,
+      "loss": 3.9554,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2018481186978464e-05,
+      "loss": 3.973,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2010095239467944e-05,
+      "loss": 3.9633,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2001709291957424e-05,
+      "loss": 3.9581,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1993323344446904e-05,
+      "loss": 3.9635,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.198493739693638e-05,
+      "loss": 3.9661,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.197655144942586e-05,
+      "loss": 3.9486,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196816550191534e-05,
+      "loss": 3.949,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.195977955440482e-05,
+      "loss": 3.9751,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1951409985698026e-05,
+      "loss": 3.965,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.194304041699124e-05,
+      "loss": 3.9618,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.193465446948072e-05,
+      "loss": 3.9483,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19262685219702e-05,
+      "loss": 3.958,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191788257445968e-05,
+      "loss": 3.9638,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190949662694916e-05,
+      "loss": 3.969,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190111067943864e-05,
+      "loss": 3.9685,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189274111073185e-05,
+      "loss": 3.9678,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188435516322133e-05,
+      "loss": 3.9687,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187596921571081e-05,
+      "loss": 3.9617,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186758326820029e-05,
+      "loss": 3.9601,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185919732068977e-05,
+      "loss": 3.9591,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185081137317925e-05,
+      "loss": 3.9645,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.184242542566873e-05,
+      "loss": 3.9627,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.183403947815821e-05,
+      "loss": 3.9565,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182565353064769e-05,
+      "loss": 3.9643,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18172839619409e-05,
+      "loss": 3.9703,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1808914393234115e-05,
+      "loss": 3.9627,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1800544824527324e-05,
+      "loss": 3.9503,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1792158877016804e-05,
+      "loss": 3.9572,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1783772929506284e-05,
+      "loss": 3.9649,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1775386981995764e-05,
+      "loss": 3.9733,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1767001034485244e-05,
+      "loss": 3.9611,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.175863146577845e-05,
+      "loss": 3.9615,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.175024551826793e-05,
+      "loss": 3.9654,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174185957075741e-05,
+      "loss": 3.9565,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173347362324689e-05,
+      "loss": 3.9577,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172508767573637e-05,
+      "loss": 3.9611,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.171670172822585e-05,
+      "loss": 3.9656,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170831578071533e-05,
+      "loss": 3.9699,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169992983320482e-05,
+      "loss": 3.9675,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16915438856943e-05,
+      "loss": 3.951,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168315793818378e-05,
+      "loss": 3.957,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.167478836947699e-05,
+      "loss": 3.9611,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16664188007702e-05,
+      "loss": 3.9692,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.165803285325968e-05,
+      "loss": 3.9564,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164964690574916e-05,
+      "loss": 3.9565,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164126095823864e-05,
+      "loss": 3.9505,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.163287501072812e-05,
+      "loss": 3.9628,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16244890632176e-05,
+      "loss": 3.9414,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161610311570708e-05,
+      "loss": 3.956,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160771716819656e-05,
+      "loss": 3.9611,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159933122068604e-05,
+      "loss": 3.9595,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159094527317552e-05,
+      "loss": 3.9535,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1582559325665e-05,
+      "loss": 3.9602,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157417337815448e-05,
+      "loss": 3.9514,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1565803809447687e-05,
+      "loss": 3.9612,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15574342407409e-05,
+      "loss": 3.9647,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154904829323038e-05,
+      "loss": 3.952,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154066234571986e-05,
+      "loss": 3.9633,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1532276398209335e-05,
+      "loss": 3.9702,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1523890450698815e-05,
+      "loss": 3.9633,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1515504503188295e-05,
+      "loss": 3.9515,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1507118555677775e-05,
+      "loss": 3.9637,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1498732608167255e-05,
+      "loss": 3.9601,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149037941826421e-05,
+      "loss": 3.9552,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.148199347075369e-05,
+      "loss": 3.9634,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.147360752324316e-05,
+      "loss": 3.956,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.146522157573264e-05,
+      "loss": 3.9624,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.145683562822212e-05,
+      "loss": 3.9512,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14484496807116e-05,
+      "loss": 3.9548,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144006373320108e-05,
+      "loss": 3.9491,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.143167778569056e-05,
+      "loss": 3.9565,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.142329183818004e-05,
+      "loss": 3.9451,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.141492226947325e-05,
+      "loss": 3.9698,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140653632196273e-05,
+      "loss": 3.9704,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139815037445221e-05,
+      "loss": 3.9539,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1389764426941696e-05,
+      "loss": 3.9587,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1381378479431176e-05,
+      "loss": 3.9646,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1373008910724385e-05,
+      "loss": 3.961,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1364622963213865e-05,
+      "loss": 3.9545,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1356237015703345e-05,
+      "loss": 3.9541,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1347867446996554e-05,
+      "loss": 3.9636,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1339481499486034e-05,
+      "loss": 3.9639,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1331095551975514e-05,
+      "loss": 3.9535,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1322709604464994e-05,
+      "loss": 3.9591,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1314323656954474e-05,
+      "loss": 3.9679,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1305937709443954e-05,
+      "loss": 3.9526,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129756814073716e-05,
+      "loss": 3.9547,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128918219322665e-05,
+      "loss": 3.9655,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128079624571613e-05,
+      "loss": 3.9538,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127241029820561e-05,
+      "loss": 3.963,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126402435069509e-05,
+      "loss": 3.9628,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.042651176452637,
+      "eval_runtime": 356.4098,
+      "eval_samples_per_second": 1070.652,
+      "eval_steps_per_second": 33.459,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125563840318457e-05,
+      "loss": 3.9399,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124725245567405e-05,
+      "loss": 3.9477,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123886650816352e-05,
+      "loss": 3.9548,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1230480560653e-05,
+      "loss": 3.9667,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122209461314248e-05,
+      "loss": 3.9623,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121370866563196e-05,
+      "loss": 3.9672,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120532271812144e-05,
+      "loss": 3.952,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119693677061092e-05,
+      "loss": 3.9619,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118856720190413e-05,
+      "loss": 3.9468,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118018125439361e-05,
+      "loss": 3.9626,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11717953068831e-05,
+      "loss": 3.953,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116340935937258e-05,
+      "loss": 3.9679,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.115502341186206e-05,
+      "loss": 3.9692,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114663746435154e-05,
+      "loss": 3.9453,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113825151684102e-05,
+      "loss": 3.9499,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11298655693305e-05,
+      "loss": 3.9415,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112147962181998e-05,
+      "loss": 3.9502,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111309367430946e-05,
+      "loss": 3.9588,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110470772679894e-05,
+      "loss": 3.9566,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109632177928841e-05,
+      "loss": 3.9515,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108793583177789e-05,
+      "loss": 3.9812,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107954988426737e-05,
+      "loss": 3.9641,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107116393675685e-05,
+      "loss": 3.9616,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.106277798924633e-05,
+      "loss": 3.9543,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1054408420539546e-05,
+      "loss": 3.9717,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1046022473029026e-05,
+      "loss": 3.9565,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1037636525518506e-05,
+      "loss": 3.954,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1029250578007986e-05,
+      "loss": 3.9641,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1020864630497466e-05,
+      "loss": 3.9456,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1012478682986946e-05,
+      "loss": 3.958,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1004109114280155e-05,
+      "loss": 3.9525,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0995723166769635e-05,
+      "loss": 3.9546,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0987337219259115e-05,
+      "loss": 3.9596,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0978951271748595e-05,
+      "loss": 3.9649,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0970565324238075e-05,
+      "loss": 3.954,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0962195755531284e-05,
+      "loss": 3.9667,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0953809808020764e-05,
+      "loss": 3.9612,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094542386051025e-05,
+      "loss": 3.9371,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093703791299973e-05,
+      "loss": 3.9522,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092865196548921e-05,
+      "loss": 3.9518,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092028239678242e-05,
+      "loss": 3.9387,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09118964492719e-05,
+      "loss": 3.9591,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.090351050176138e-05,
+      "loss": 3.9605,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089512455425086e-05,
+      "loss": 3.9565,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088673860674034e-05,
+      "loss": 3.9487,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087835265922982e-05,
+      "loss": 3.9633,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08699667117193e-05,
+      "loss": 3.9504,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086158076420877e-05,
+      "loss": 3.9686,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085319481669825e-05,
+      "loss": 3.938,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084482524799147e-05,
+      "loss": 3.956,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.083643930048095e-05,
+      "loss": 3.9597,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082805335297043e-05,
+      "loss": 3.9515,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081966740545991e-05,
+      "loss": 3.9573,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0811314215556854e-05,
+      "loss": 3.9387,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0802928268046333e-05,
+      "loss": 3.9551,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0794542320535813e-05,
+      "loss": 3.937,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0786156373025293e-05,
+      "loss": 3.9665,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07777868043185e-05,
+      "loss": 3.9462,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076940085680798e-05,
+      "loss": 3.9567,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076101490929746e-05,
+      "loss": 3.955,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075262896178694e-05,
+      "loss": 3.9453,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074424301427642e-05,
+      "loss": 3.9501,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073587344556964e-05,
+      "loss": 3.9537,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072748749805912e-05,
+      "loss": 3.9361,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07191015505486e-05,
+      "loss": 3.9374,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071071560303807e-05,
+      "loss": 3.9589,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070232965552755e-05,
+      "loss": 3.9526,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069394370801703e-05,
+      "loss": 3.9525,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.068555776050651e-05,
+      "loss": 3.936,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067717181299599e-05,
+      "loss": 3.9433,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06688022442892e-05,
+      "loss": 3.9497,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066041629677868e-05,
+      "loss": 3.9563,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.065203034926816e-05,
+      "loss": 3.9578,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.064364440175764e-05,
+      "loss": 3.9525,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.063525845424712e-05,
+      "loss": 3.9593,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0626888885540336e-05,
+      "loss": 3.9489,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0618502938029816e-05,
+      "loss": 3.9452,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0610116990519296e-05,
+      "loss": 3.948,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0601747421812505e-05,
+      "loss": 3.9473,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0593361474301985e-05,
+      "loss": 3.9507,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0584975526791465e-05,
+      "loss": 3.9467,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0576589579280945e-05,
+      "loss": 3.9464,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0568203631770425e-05,
+      "loss": 3.9663,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0559834063063634e-05,
+      "loss": 3.9482,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0551448115553114e-05,
+      "loss": 3.9391,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0543062168042594e-05,
+      "loss": 3.9446,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0534676220532074e-05,
+      "loss": 3.9508,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052630665182529e-05,
+      "loss": 3.9572,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051792070431477e-05,
+      "loss": 3.9496,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.050953475680425e-05,
+      "loss": 3.9468,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.050114880929373e-05,
+      "loss": 3.9542,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.049276286178321e-05,
+      "loss": 3.9462,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.048439329307642e-05,
+      "loss": 3.9447,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.04760073455659e-05,
+      "loss": 3.9491,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.046762139805538e-05,
+      "loss": 3.9529,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.045923545054486e-05,
+      "loss": 3.9592,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0450849503034338e-05,
+      "loss": 3.9525,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.044246355552382e-05,
+      "loss": 3.9426,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.043409398681703e-05,
+      "loss": 3.9429,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.042570803930651e-05,
+      "loss": 3.9477,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041732209179599e-05,
+      "loss": 3.9601,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040893614428547e-05,
+      "loss": 3.9442,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040055019677495e-05,
+      "loss": 3.9438,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.039216424926443e-05,
+      "loss": 3.9411,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0383778301753907e-05,
+      "loss": 3.9477,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375408733047123e-05,
+      "loss": 3.9297,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0367022785536603e-05,
+      "loss": 3.9471,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0358636838026076e-05,
+      "loss": 3.9452,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035025089051556e-05,
+      "loss": 3.9492,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0341881321808775e-05,
+      "loss": 3.9417,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0333495374298255e-05,
+      "loss": 3.9513,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032510942678773e-05,
+      "loss": 3.9383,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.031672347927721e-05,
+      "loss": 3.9488,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.030833753176669e-05,
+      "loss": 3.9531,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.029995158425617e-05,
+      "loss": 3.9389,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.029156563674565e-05,
+      "loss": 3.9536,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.028319606803886e-05,
+      "loss": 3.9578,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0274826499332077e-05,
+      "loss": 3.9529,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.026644055182155e-05,
+      "loss": 3.9426,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.025805460431103e-05,
+      "loss": 3.9445,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0249668656800513e-05,
+      "loss": 3.9512,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0241282709289993e-05,
+      "loss": 3.9429,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0232896761779473e-05,
+      "loss": 3.9527,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0224510814268953e-05,
+      "loss": 3.9441,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0216124866758433e-05,
+      "loss": 3.9536,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0207738919247913e-05,
+      "loss": 3.9382,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0199352971737393e-05,
+      "loss": 3.9434,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0190967024226873e-05,
+      "loss": 3.9367,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182597455520085e-05,
+      "loss": 3.9431,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174211508009565e-05,
+      "loss": 3.9333,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0165825560499045e-05,
+      "loss": 3.9567,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0157455991792254e-05,
+      "loss": 3.9637,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0149070044281734e-05,
+      "loss": 3.9442,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0140684096771214e-05,
+      "loss": 3.9426,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0132298149260698e-05,
+      "loss": 3.9529,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0123912201750177e-05,
+      "loss": 3.954,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0115526254239657e-05,
+      "loss": 3.9453,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0107140306729137e-05,
+      "loss": 3.9401,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0098754359218617e-05,
+      "loss": 3.9556,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009036841170809e-05,
+      "loss": 3.9497,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0081998843001306e-05,
+      "loss": 3.9457,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.007361289549079e-05,
+      "loss": 3.945,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.006522694798027e-05,
+      "loss": 3.9542,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0056841000469743e-05,
+      "loss": 3.9424,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004847143176296e-05,
+      "loss": 3.9435,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004008548425244e-05,
+      "loss": 3.9548,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0031699536741915e-05,
+      "loss": 3.9447,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023313589231395e-05,
+      "loss": 3.9488,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0014927641720875e-05,
+      "loss": 3.9543,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0006541694210355e-05,
+      "loss": 3.946,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.036015033721924,
+      "eval_runtime": 286.3134,
+      "eval_samples_per_second": 1332.774,
+      "eval_steps_per_second": 41.65,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998155746699835e-05,
+      "loss": 3.9368,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989769799189315e-05,
+      "loss": 3.9398,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981383851678795e-05,
+      "loss": 3.9538,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9972997904168275e-05,
+      "loss": 3.9507,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964611956657755e-05,
+      "loss": 3.9538,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956226009147238e-05,
+      "loss": 3.9402,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9947840061636718e-05,
+      "loss": 3.9531,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939454114126198e-05,
+      "loss": 3.9347,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931068166615678e-05,
+      "loss": 3.9536,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.992268221910515e-05,
+      "loss": 3.9385,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.991429627159463e-05,
+      "loss": 3.958,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9905910324084115e-05,
+      "loss": 3.9542,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.989754075537733e-05,
+      "loss": 3.9387,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889154807866804e-05,
+      "loss": 3.9384,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880768860356284e-05,
+      "loss": 3.9324,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9872382912845764e-05,
+      "loss": 3.9359,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9863996965335244e-05,
+      "loss": 3.9495,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9855611017824723e-05,
+      "loss": 3.944,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9847225070314207e-05,
+      "loss": 3.9435,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9838839122803687e-05,
+      "loss": 3.968,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830469554096896e-05,
+      "loss": 3.9506,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822083606586376e-05,
+      "loss": 3.9516,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813697659075856e-05,
+      "loss": 3.9416,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9805311711565336e-05,
+      "loss": 3.9584,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9796942142858548e-05,
+      "loss": 3.9413,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788556195348028e-05,
+      "loss": 3.9499,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9780170247837508e-05,
+      "loss": 3.9496,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9771784300326988e-05,
+      "loss": 3.9401,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763398352816468e-05,
+      "loss": 3.9422,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755028784109677e-05,
+      "loss": 3.9416,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.974664283659916e-05,
+      "loss": 3.9448,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973825688908864e-05,
+      "loss": 3.9479,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972987094157812e-05,
+      "loss": 3.954,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.97214849940676e-05,
+      "loss": 3.9483,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.971311542536081e-05,
+      "loss": 3.9514,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.970472947785029e-05,
+      "loss": 3.9497,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.969634353033977e-05,
+      "loss": 3.9308,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9687957582829253e-05,
+      "loss": 3.9414,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9679571635318733e-05,
+      "loss": 3.9386,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9671185687808213e-05,
+      "loss": 3.9303,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9662799740297693e-05,
+      "loss": 3.9475,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9654413792787166e-05,
+      "loss": 3.952,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.964604422408038e-05,
+      "loss": 3.9422,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963765827656986e-05,
+      "loss": 3.9407,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9629272329059345e-05,
+      "loss": 3.9505,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9620886381548818e-05,
+      "loss": 3.9446,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9612533191645763e-05,
+      "loss": 3.9509,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9604147244135243e-05,
+      "loss": 3.9351,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9595761296624723e-05,
+      "loss": 3.9407,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9587375349114206e-05,
+      "loss": 3.9504,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9578989401603686e-05,
+      "loss": 3.9422,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9570619832896895e-05,
+      "loss": 3.9445,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562233885386375e-05,
+      "loss": 3.9343,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9553847937875855e-05,
+      "loss": 3.945,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545461990365335e-05,
+      "loss": 3.927,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537076042854815e-05,
+      "loss": 3.9542,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528690095344292e-05,
+      "loss": 3.9386,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520304147833772e-05,
+      "loss": 3.9441,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9511918200323252e-05,
+      "loss": 3.9411,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.950354863161646e-05,
+      "loss": 3.9388,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9495162684105944e-05,
+      "loss": 3.9364,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486776736595424e-05,
+      "loss": 3.9451,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478390789084904e-05,
+      "loss": 3.9285,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9470004841574384e-05,
+      "loss": 3.9275,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9461635272867593e-05,
+      "loss": 3.9464,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9453249325357073e-05,
+      "loss": 3.947,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9444863377846553e-05,
+      "loss": 3.941,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9436477430336033e-05,
+      "loss": 3.9264,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942812424043298e-05,
+      "loss": 3.9335,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941973829292246e-05,
+      "loss": 3.94,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411352345411935e-05,
+      "loss": 3.9469,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9402966397901415e-05,
+      "loss": 3.9443,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394580450390894e-05,
+      "loss": 3.9439,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.938621088168411e-05,
+      "loss": 3.9493,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377824934173587e-05,
+      "loss": 3.9464,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369438986663067e-05,
+      "loss": 3.9308,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9361053039152547e-05,
+      "loss": 3.9372,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9352667091642027e-05,
+      "loss": 3.9445,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344281144131507e-05,
+      "loss": 3.9373,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.933591157542472e-05,
+      "loss": 3.9384,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.93275256279142e-05,
+      "loss": 3.9359,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931913968040368e-05,
+      "loss": 3.9552,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310770111696888e-05,
+      "loss": 3.9367,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302384164186368e-05,
+      "loss": 3.9275,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9293998216675848e-05,
+      "loss": 3.9373,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.928561226916533e-05,
+      "loss": 3.942,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.927722632165481e-05,
+      "loss": 3.9483,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.926884037414429e-05,
+      "loss": 3.9361,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.926045442663377e-05,
+      "loss": 3.9383,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.925206847912325e-05,
+      "loss": 3.9415,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.924369891041646e-05,
+      "loss": 3.9384,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.923531296290594e-05,
+      "loss": 3.9317,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9226927015395424e-05,
+      "loss": 3.9377,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9218541067884904e-05,
+      "loss": 3.9452,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9210155120374384e-05,
+      "loss": 3.9537,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9201769172863864e-05,
+      "loss": 3.9358,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9193383225353344e-05,
+      "loss": 3.9321,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9185030035450285e-05,
+      "loss": 3.9386,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9176644087939765e-05,
+      "loss": 3.9388,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9168258140429245e-05,
+      "loss": 3.9473,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9159872192918725e-05,
+      "loss": 3.9361,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9151486245408205e-05,
+      "loss": 3.9348,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143100297897685e-05,
+      "loss": 3.9301,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9134714350387165e-05,
+      "loss": 3.9433,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126328402876645e-05,
+      "loss": 3.9162,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.911794245536612e-05,
+      "loss": 3.9369,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91095565078556e-05,
+      "loss": 3.9341,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910117056034508e-05,
+      "loss": 3.9403,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092800991638297e-05,
+      "loss": 3.9317,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.908441504412777e-05,
+      "loss": 3.942,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076045475420986e-05,
+      "loss": 3.9308,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.90676759067142e-05,
+      "loss": 3.9354,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905928995920368e-05,
+      "loss": 3.9401,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905090401169316e-05,
+      "loss": 3.9318,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904251806418264e-05,
+      "loss": 3.9468,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.903413211667212e-05,
+      "loss": 3.9496,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025746169161595e-05,
+      "loss": 3.9415,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9017360221651075e-05,
+      "loss": 3.9345,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9008974274140555e-05,
+      "loss": 3.9343,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000588326630035e-05,
+      "loss": 3.9391,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992202379119515e-05,
+      "loss": 3.9345,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8983816431608995e-05,
+      "loss": 3.9422,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8975430484098475e-05,
+      "loss": 3.9375,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8967044536587955e-05,
+      "loss": 3.9407,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958658589077438e-05,
+      "loss": 3.9299,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950272641566918e-05,
+      "loss": 3.9338,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8941886694056398e-05,
+      "loss": 3.9258,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933533504153336e-05,
+      "loss": 3.9313,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925147556642816e-05,
+      "loss": 3.922,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.89167616091323e-05,
+      "loss": 3.9463,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.890837566162178e-05,
+      "loss": 3.955,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.889998971411126e-05,
+      "loss": 3.9426,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.889160376660074e-05,
+      "loss": 3.925,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.888323419789395e-05,
+      "loss": 3.9454,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887484825038343e-05,
+      "loss": 3.9419,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886646230287291e-05,
+      "loss": 3.9358,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.885807635536239e-05,
+      "loss": 3.9303,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849690407851872e-05,
+      "loss": 3.9441,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841304460341352e-05,
+      "loss": 3.9381,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.883293489163456e-05,
+      "loss": 3.9346,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882454894412404e-05,
+      "loss": 3.9387,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881616299661352e-05,
+      "loss": 3.9432,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807777049103e-05,
+      "loss": 3.934,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879939110159248e-05,
+      "loss": 3.932,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791005154081957e-05,
+      "loss": 3.948,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8782635585375173e-05,
+      "loss": 3.9339,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8774266016668382e-05,
+      "loss": 3.9375,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8765880069157862e-05,
+      "loss": 3.9456,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8757494121647342e-05,
+      "loss": 3.9308,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.031096458435059,
+      "eval_runtime": 299.0363,
+      "eval_samples_per_second": 1276.069,
+      "eval_steps_per_second": 39.878,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8749108174136825e-05,
+      "loss": 3.9278,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8740722226626305e-05,
+      "loss": 3.9343,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.873233627911578e-05,
+      "loss": 3.9435,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.872395033160526e-05,
+      "loss": 3.9393,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.871556438409474e-05,
+      "loss": 3.9465,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.870717843658422e-05,
+      "loss": 3.9302,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.869880886787743e-05,
+      "loss": 3.9394,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.869042292036691e-05,
+      "loss": 3.9261,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.868203697285639e-05,
+      "loss": 3.9439,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.867365102534587e-05,
+      "loss": 3.9313,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.866526507783535e-05,
+      "loss": 3.945,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.865687913032483e-05,
+      "loss": 3.9419,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8648509561618043e-05,
+      "loss": 3.9344,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640123614107523e-05,
+      "loss": 3.9284,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631737666597003e-05,
+      "loss": 3.9251,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8623351719086483e-05,
+      "loss": 3.9261,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8614982150379692e-05,
+      "loss": 3.9375,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8606612581672905e-05,
+      "loss": 3.9364,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8598226634162385e-05,
+      "loss": 3.9322,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8589840686651865e-05,
+      "loss": 3.9568,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8581454739141344e-05,
+      "loss": 3.9467,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8573068791630824e-05,
+      "loss": 3.9405,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564682844120304e-05,
+      "loss": 3.9329,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8556296896609784e-05,
+      "loss": 3.9514,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8547910949099264e-05,
+      "loss": 3.929,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539525001588748e-05,
+      "loss": 3.9389,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531139054078228e-05,
+      "loss": 3.939,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8522753106567708e-05,
+      "loss": 3.9351,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8514367159057188e-05,
+      "loss": 3.9297,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8505997590350397e-05,
+      "loss": 3.9331,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.849762802164361e-05,
+      "loss": 3.9352,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.848924207413309e-05,
+      "loss": 3.9372,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.848085612662257e-05,
+      "loss": 3.9411,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.847247017911205e-05,
+      "loss": 3.9413,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464100610405258e-05,
+      "loss": 3.9429,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455714662894738e-05,
+      "loss": 3.937,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447328715384218e-05,
+      "loss": 3.923,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.84389427678737e-05,
+      "loss": 3.9356,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.843055682036318e-05,
+      "loss": 3.9281,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.842217087285266e-05,
+      "loss": 3.9244,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.841378492534214e-05,
+      "loss": 3.9384,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8405398977831614e-05,
+      "loss": 3.9421,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8397013030321094e-05,
+      "loss": 3.9308,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8388627082810578e-05,
+      "loss": 3.9313,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8380241135300058e-05,
+      "loss": 3.9377,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8371855187789538e-05,
+      "loss": 3.9365,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8363518376690212e-05,
+      "loss": 3.9402,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.835513242917969e-05,
+      "loss": 3.9301,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.834674648166917e-05,
+      "loss": 3.9311,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8338360534158655e-05,
+      "loss": 3.938,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8329974586648135e-05,
+      "loss": 3.9308,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8321588639137615e-05,
+      "loss": 3.9327,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8313202691627088e-05,
+      "loss": 3.9275,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304816744116568e-05,
+      "loss": 3.937,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296430796606048e-05,
+      "loss": 3.9153,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8288061227899264e-05,
+      "loss": 3.9403,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.827967528038874e-05,
+      "loss": 3.9295,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.827128933287822e-05,
+      "loss": 3.9306,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.82629033853677e-05,
+      "loss": 3.9386,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.825451743785718e-05,
+      "loss": 3.9277,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.824613149034666e-05,
+      "loss": 3.9243,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.823774554283614e-05,
+      "loss": 3.9405,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.822935959532562e-05,
+      "loss": 3.9145,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8220990026618833e-05,
+      "loss": 3.9196,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212604079108313e-05,
+      "loss": 3.9351,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8204218131597793e-05,
+      "loss": 3.9372,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8195848562891e-05,
+      "loss": 3.9309,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.818746261538048e-05,
+      "loss": 3.9171,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8179076667869965e-05,
+      "loss": 3.9222,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8170690720359445e-05,
+      "loss": 3.9269,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8162304772848925e-05,
+      "loss": 3.9392,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8153918825338405e-05,
+      "loss": 3.9325,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8145532877827885e-05,
+      "loss": 3.9371,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8137146930317365e-05,
+      "loss": 3.9369,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128777361610574e-05,
+      "loss": 3.9399,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8120391414100057e-05,
+      "loss": 3.9199,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8112005466589537e-05,
+      "loss": 3.9287,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8103619519079017e-05,
+      "loss": 3.9379,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8095249950372226e-05,
+      "loss": 3.9266,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8086864002861706e-05,
+      "loss": 3.9333,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.807849443415492e-05,
+      "loss": 3.9221,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.80701084866444e-05,
+      "loss": 3.9466,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.806172253913388e-05,
+      "loss": 3.9285,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.805333659162336e-05,
+      "loss": 3.9186,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.804495064411284e-05,
+      "loss": 3.9263,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.803656469660232e-05,
+      "loss": 3.9325,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.80281787490918e-05,
+      "loss": 3.9383,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801980918038501e-05,
+      "loss": 3.9276,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801142323287449e-05,
+      "loss": 3.9288,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.800303728536397e-05,
+      "loss": 3.9349,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.799465133785345e-05,
+      "loss": 3.9269,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.798628176914666e-05,
+      "loss": 3.9225,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.797789582163614e-05,
+      "loss": 3.9267,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796950987412562e-05,
+      "loss": 3.9391,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7961123926615103e-05,
+      "loss": 3.9371,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7952737979104576e-05,
+      "loss": 3.9338,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7944368410397792e-05,
+      "loss": 3.9207,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7935982462887272e-05,
+      "loss": 3.9265,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.792759651537675e-05,
+      "loss": 3.9308,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919226946669965e-05,
+      "loss": 3.9335,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7910840999159444e-05,
+      "loss": 3.9298,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7902455051648924e-05,
+      "loss": 3.9268,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7894069104138398e-05,
+      "loss": 3.9196,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885683156627878e-05,
+      "loss": 3.9348,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877297209117358e-05,
+      "loss": 3.9041,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786891126160684e-05,
+      "loss": 3.9312,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786052531409632e-05,
+      "loss": 3.9218,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.78521393665858e-05,
+      "loss": 3.9327,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.784375341907528e-05,
+      "loss": 3.9198,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.783536747156476e-05,
+      "loss": 3.9334,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.782699790285797e-05,
+      "loss": 3.9207,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.781861195534745e-05,
+      "loss": 3.9248,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810226007836933e-05,
+      "loss": 3.9327,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7801840060326413e-05,
+      "loss": 3.9237,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7793454112815893e-05,
+      "loss": 3.9346,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7785068165305373e-05,
+      "loss": 3.9397,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7776698596598582e-05,
+      "loss": 3.931,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768312649088062e-05,
+      "loss": 3.9235,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7759926701577542e-05,
+      "loss": 3.9271,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7751540754067022e-05,
+      "loss": 3.9313,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7743154806556505e-05,
+      "loss": 3.9191,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734768859045985e-05,
+      "loss": 3.9371,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726382911535465e-05,
+      "loss": 3.9235,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.771799696402494e-05,
+      "loss": 3.9325,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7709627395318154e-05,
+      "loss": 3.9263,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7701241447807634e-05,
+      "loss": 3.9223,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.769285550029711e-05,
+      "loss": 3.9141,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.768446955278659e-05,
+      "loss": 3.9241,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.767608360527607e-05,
+      "loss": 3.9147,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667714036569287e-05,
+      "loss": 3.9374,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765932808905876e-05,
+      "loss": 3.9481,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7650942141548243e-05,
+      "loss": 3.9318,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7642556194037723e-05,
+      "loss": 3.9165,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634170246527203e-05,
+      "loss": 3.9338,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7625784299016683e-05,
+      "loss": 3.9279,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7617398351506163e-05,
+      "loss": 3.9288,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7609012403995643e-05,
+      "loss": 3.9268,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7600659214092584e-05,
+      "loss": 3.9309,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7592273266582064e-05,
+      "loss": 3.9257,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7583887319071544e-05,
+      "loss": 3.9285,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7575501371561024e-05,
+      "loss": 3.9302,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7567131802854233e-05,
+      "loss": 3.9353,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7558745855343713e-05,
+      "loss": 3.9292,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755037628663693e-05,
+      "loss": 3.9198,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7541990339126406e-05,
+      "loss": 3.9383,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7533604391615886e-05,
+      "loss": 3.9294,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7525218444105366e-05,
+      "loss": 3.9238,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.751684887539858e-05,
+      "loss": 3.9387,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508462927888058e-05,
+      "loss": 3.9241,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.026576519012451,
+      "eval_runtime": 298.136,
+      "eval_samples_per_second": 1279.922,
+      "eval_steps_per_second": 39.999,
+      "step": 1373760
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 5.5007188738557126e+17,
+  "trial_name": null,
+  "trial_params": null
+}