diff --git "a/checkpoint-534240/trainer_state.json" "b/checkpoint-534240/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-534240/trainer_state.json"
@@ -0,0 +1,6339 @@
+{
+  "best_metric": 4.151233673095703,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/lstm/3/checkpoints/checkpoint-534240",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 534240,
+  "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
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.1391684509438883e+17,
+  "trial_name": null,
+  "trial_params": null
+}