diff --git "a/checkpoint-610552/trainer_state.json" "b/checkpoint-610552/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-610552/trainer_state.json"
@@ -0,0 +1,7241 @@
+{
+  "best_metric": 4.107790470123291,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/lstm/4/checkpoints/checkpoint-610552",
+  "epoch": 0.025000278439663435,
+  "eval_steps": 10,
+  "global_step": 610552,
+  "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": 10.8202,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5642,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0561,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9848,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9516,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8317,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7218,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6162,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5336,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4403,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3866,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3152,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2499,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.185,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.135,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0919,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986585759743914e-05,
+      "loss": 6.0468,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985747164992862e-05,
+      "loss": 5.9949,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98490857024181e-05,
+      "loss": 5.96,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984069975490758e-05,
+      "loss": 5.926,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.881,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8466,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8268,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7903,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7694,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7434,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7206,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6927,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6587,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6528,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6204,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6044,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5801,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5763,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5452,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5375,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5231,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.5038,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 5.4937,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.4722,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4475,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4423,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4142,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.4262,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3996,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3943,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3795,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3756,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3695,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3395,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3284,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3211,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3282,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.3042,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.3108,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.2807,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2767,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.264,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2578,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.255,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.2339,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2284,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2327,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2108,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.2101,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.198,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.198,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.182,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1854,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 5.1715,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.162,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1647,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1511,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1585,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1559,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 5.1459,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1271,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1115,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.1044,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.1102,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 5.1134,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 5.1064,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.0816,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0762,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.085,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0699,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.0698,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.0613,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.063,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0524,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 5.0435,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 5.0299,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0439,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 5.0289,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 5.0127,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 5.0113,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 5.0246,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 5.0149,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 5.0059,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9942,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.998,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.9769,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.9922,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.9944,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.9695,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9626,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.94,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9518,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9525,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9467,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9475,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9316,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9326,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.9338,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9357,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9228,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.9119,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9132,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.9165,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.9104,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9059,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8968,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8881,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8811,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8877,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8874,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8881,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8623,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892694267353179e-05,
+      "loss": 4.876,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8748,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8618,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.8774,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8571,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.8708,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.858,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8387,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.8374,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8375,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843115956034054e-05,
+      "loss": 4.8411,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834730008523534e-05,
+      "loss": 4.8453,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826344061013014e-05,
+      "loss": 4.8415,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8817958113502494e-05,
+      "loss": 4.8348,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.8253,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8133,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8126,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.8184,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.8269,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.8172,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.8004,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87509032910258e-05,
+      "loss": 4.8004,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.7640604972839355,
+      "eval_runtime": 557.5755,
+      "eval_samples_per_second": 684.376,
+      "eval_steps_per_second": 21.387,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874251734351528e-05,
+      "loss": 4.8081,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873413139600475e-05,
+      "loss": 4.8005,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872574544849423e-05,
+      "loss": 4.7917,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871735950098371e-05,
+      "loss": 4.7901,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870897355347319e-05,
+      "loss": 4.7813,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870058760596267e-05,
+      "loss": 4.7651,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869220165845215e-05,
+      "loss": 4.7739,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.868381571094163e-05,
+      "loss": 4.7833,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867542976343111e-05,
+      "loss": 4.7657,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866704381592059e-05,
+      "loss": 4.774,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865865786841007e-05,
+      "loss": 4.7773,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865027192089955e-05,
+      "loss": 4.7754,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864188597338903e-05,
+      "loss": 4.7656,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863350002587852e-05,
+      "loss": 4.7627,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8625114078368e-05,
+      "loss": 4.7588,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861672813085747e-05,
+      "loss": 4.7492,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860835856215069e-05,
+      "loss": 4.7519,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.859997261464017e-05,
+      "loss": 4.7529,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8591603045933376e-05,
+      "loss": 4.7516,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8583217098422856e-05,
+      "loss": 4.7499,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574831150912336e-05,
+      "loss": 4.7374,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566445203401816e-05,
+      "loss": 4.7492,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.855805925589129e-05,
+      "loss": 4.7341,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8549689687184505e-05,
+      "loss": 4.7438,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541303739673985e-05,
+      "loss": 4.7327,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8532917792163465e-05,
+      "loss": 4.7319,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8524531844652945e-05,
+      "loss": 4.7353,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8516145897142425e-05,
+      "loss": 4.7153,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850777632843564e-05,
+      "loss": 4.7304,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8499390380925114e-05,
+      "loss": 4.7108,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8491004433414594e-05,
+      "loss": 4.7166,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8482618485904074e-05,
+      "loss": 4.712,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8474232538393554e-05,
+      "loss": 4.7209,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8465846590883034e-05,
+      "loss": 4.7075,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8457460643372514e-05,
+      "loss": 4.7017,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8449074695861993e-05,
+      "loss": 4.7185,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8440688748351473e-05,
+      "loss": 4.7052,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843231917964469e-05,
+      "loss": 4.711,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.842393323213417e-05,
+      "loss": 4.6993,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841554728462365e-05,
+      "loss": 4.6847,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840716133711313e-05,
+      "loss": 4.6887,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839877538960261e-05,
+      "loss": 4.6801,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839038944209209e-05,
+      "loss": 4.6946,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83820198733853e-05,
+      "loss": 4.6843,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837363392587478e-05,
+      "loss": 4.6909,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836524797836426e-05,
+      "loss": 4.6826,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835686203085374e-05,
+      "loss": 4.6879,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834847608334322e-05,
+      "loss": 4.6849,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834010651463643e-05,
+      "loss": 4.6609,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833173694592964e-05,
+      "loss": 4.6653,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832335099841912e-05,
+      "loss": 4.6635,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83149650509086e-05,
+      "loss": 4.6761,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830657910339808e-05,
+      "loss": 4.6551,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829820953469129e-05,
+      "loss": 4.6784,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828982358718077e-05,
+      "loss": 4.6518,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828143763967025e-05,
+      "loss": 4.6514,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.827305169215973e-05,
+      "loss": 4.6557,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.826466574464921e-05,
+      "loss": 4.6498,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.825627979713869e-05,
+      "loss": 4.651,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.824789384962817e-05,
+      "loss": 4.644,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823950790211765e-05,
+      "loss": 4.6407,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231121954607125e-05,
+      "loss": 4.6519,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.822273600709661e-05,
+      "loss": 4.6366,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.821435005958609e-05,
+      "loss": 4.638,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.82059804908793e-05,
+      "loss": 4.6354,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.819759454336878e-05,
+      "loss": 4.6399,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818920859585826e-05,
+      "loss": 4.6226,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818082264834774e-05,
+      "loss": 4.643,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817245307964095e-05,
+      "loss": 4.6293,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816406713213043e-05,
+      "loss": 4.6273,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.815568118461991e-05,
+      "loss": 4.6335,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814729523710939e-05,
+      "loss": 4.6205,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813890928959887e-05,
+      "loss": 4.6381,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813053972089208e-05,
+      "loss": 4.6429,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8122153773381565e-05,
+      "loss": 4.6364,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8113784204674774e-05,
+      "loss": 4.6166,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8105398257164254e-05,
+      "loss": 4.6117,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8097012309653734e-05,
+      "loss": 4.6079,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8088626362143214e-05,
+      "loss": 4.6183,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8080240414632694e-05,
+      "loss": 4.6282,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8071854467122174e-05,
+      "loss": 4.6213,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8063468519611654e-05,
+      "loss": 4.6067,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8055082572101134e-05,
+      "loss": 4.5945,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8046696624590614e-05,
+      "loss": 4.6138,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8038310677080094e-05,
+      "loss": 4.5986,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8029924729569574e-05,
+      "loss": 4.6017,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802155516086278e-05,
+      "loss": 4.6045,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8013185592156e-05,
+      "loss": 4.6106,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800479964464548e-05,
+      "loss": 4.6002,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799641369713496e-05,
+      "loss": 4.5959,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798802774962444e-05,
+      "loss": 4.5874,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797964180211392e-05,
+      "loss": 4.5988,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.79712558546034e-05,
+      "loss": 4.5983,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796286990709288e-05,
+      "loss": 4.5736,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795448395958236e-05,
+      "loss": 4.5826,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794609801207184e-05,
+      "loss": 4.5981,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793772844336505e-05,
+      "loss": 4.5993,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792934249585453e-05,
+      "loss": 4.5856,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792095654834401e-05,
+      "loss": 4.581,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791257060083349e-05,
+      "loss": 4.5896,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7904201032126697e-05,
+      "loss": 4.5702,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789581508461618e-05,
+      "loss": 4.5854,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788742913710566e-05,
+      "loss": 4.5887,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7879043189595136e-05,
+      "loss": 4.5766,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7870657242084616e-05,
+      "loss": 4.5751,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7862271294574096e-05,
+      "loss": 4.5525,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7853885347063576e-05,
+      "loss": 4.559,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845499399553056e-05,
+      "loss": 4.5642,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.783714620965e-05,
+      "loss": 4.5675,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782877664094321e-05,
+      "loss": 4.5715,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782039069343269e-05,
+      "loss": 4.5585,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.781200474592217e-05,
+      "loss": 4.5562,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780361879841165e-05,
+      "loss": 4.5631,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.779523285090114e-05,
+      "loss": 4.57,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778684690339061e-05,
+      "loss": 4.557,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777846095588009e-05,
+      "loss": 4.5568,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777007500836957e-05,
+      "loss": 4.5543,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7761705439662786e-05,
+      "loss": 4.5576,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775331949215226e-05,
+      "loss": 4.5573,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774493354464174e-05,
+      "loss": 4.5507,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773654759713122e-05,
+      "loss": 4.5492,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77281616496207e-05,
+      "loss": 4.5444,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771977570211018e-05,
+      "loss": 4.5384,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7711422512207124e-05,
+      "loss": 4.5446,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7703036564696604e-05,
+      "loss": 4.5483,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7694650617186084e-05,
+      "loss": 4.5505,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7686264669675564e-05,
+      "loss": 4.5334,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.767789510096878e-05,
+      "loss": 4.5408,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766950915345826e-05,
+      "loss": 4.5505,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766112320594773e-05,
+      "loss": 4.5363,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.765273725843721e-05,
+      "loss": 4.5483,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764435131092669e-05,
+      "loss": 4.5348,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.763596536341617e-05,
+      "loss": 4.5522,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762759579470938e-05,
+      "loss": 4.5377,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761920984719886e-05,
+      "loss": 4.5234,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761082389968834e-05,
+      "loss": 4.5257,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760243795217783e-05,
+      "loss": 4.5236,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759405200466731e-05,
+      "loss": 4.5338,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758566605715679e-05,
+      "loss": 4.5354,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757728010964627e-05,
+      "loss": 4.5352,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756889416213575e-05,
+      "loss": 4.5297,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756052459342896e-05,
+      "loss": 4.5223,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755213864591844e-05,
+      "loss": 4.516,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754375269840792e-05,
+      "loss": 4.5138,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75353667508974e-05,
+      "loss": 4.5242,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.752698080338688e-05,
+      "loss": 4.5374,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751859485587636e-05,
+      "loss": 4.5214,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7510225287169566e-05,
+      "loss": 4.5135,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7501839339659046e-05,
+      "loss": 4.5121,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.480572700500488,
+      "eval_runtime": 532.969,
+      "eval_samples_per_second": 715.972,
+      "eval_steps_per_second": 22.375,
+      "step": 152638
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7493453392148526e-05,
+      "loss": 4.5232,
+      "step": 153088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.748506744463801e-05,
+      "loss": 4.5198,
+      "step": 153600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.747668149712749e-05,
+      "loss": 4.5057,
+      "step": 154112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.5075,
+      "step": 154624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.506,
+      "step": 155136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745155641220339e-05,
+      "loss": 4.4892,
+      "step": 155648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.744317046469287e-05,
+      "loss": 4.5005,
+      "step": 156160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.743478451718235e-05,
+      "loss": 4.5108,
+      "step": 156672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.742639856967183e-05,
+      "loss": 4.49,
+      "step": 157184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.741801262216131e-05,
+      "loss": 4.5015,
+      "step": 157696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.740962667465079e-05,
+      "loss": 4.5125,
+      "step": 158208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7401240727140264e-05,
+      "loss": 4.5087,
+      "step": 158720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.739285477962975e-05,
+      "loss": 4.4966,
+      "step": 159232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.738446883211923e-05,
+      "loss": 4.498,
+      "step": 159744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.737608288460871e-05,
+      "loss": 4.4959,
+      "step": 160256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4892,
+      "step": 160768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.4889,
+      "step": 161280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4949,
+      "step": 161792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.734255547337036e-05,
+      "loss": 4.4973,
+      "step": 162304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.733416952585984e-05,
+      "loss": 4.4949,
+      "step": 162816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.732578357834932e-05,
+      "loss": 4.4845,
+      "step": 163328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4972,
+      "step": 163840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4854,
+      "step": 164352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4937,
+      "step": 164864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4867,
+      "step": 165376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4842,
+      "step": 165888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.727551702969739e-05,
+      "loss": 4.49,
+      "step": 166400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.726713108218687e-05,
+      "loss": 4.4728,
+      "step": 166912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.725874513467635e-05,
+      "loss": 4.4873,
+      "step": 167424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.725035918716583e-05,
+      "loss": 4.4727,
+      "step": 167936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.724197323965531e-05,
+      "loss": 4.4757,
+      "step": 168448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.723358729214479e-05,
+      "loss": 4.4752,
+      "step": 168960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.722520134463427e-05,
+      "loss": 4.4794,
+      "step": 169472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.721681539712375e-05,
+      "loss": 4.474,
+      "step": 169984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4659,
+      "step": 170496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720005988090644e-05,
+      "loss": 4.4844,
+      "step": 171008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.719167393339592e-05,
+      "loss": 4.4723,
+      "step": 171520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.71832879858854e-05,
+      "loss": 4.4842,
+      "step": 172032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.717490203837488e-05,
+      "loss": 4.4713,
+      "step": 172544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.716651609086437e-05,
+      "loss": 4.4566,
+      "step": 173056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.715813014335385e-05,
+      "loss": 4.4576,
+      "step": 173568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.4588,
+      "step": 174080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714135824833281e-05,
+      "loss": 4.4779,
+      "step": 174592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4544,
+      "step": 175104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4684,
+      "step": 175616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.711621678460498e-05,
+      "loss": 4.4654,
+      "step": 176128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4626,
+      "step": 176640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7099461268387667e-05,
+      "loss": 4.4693,
+      "step": 177152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7091075320877147e-05,
+      "loss": 4.4428,
+      "step": 177664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7082689373366627e-05,
+      "loss": 4.4503,
+      "step": 178176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7074319804659836e-05,
+      "loss": 4.4428,
+      "step": 178688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.706593385714932e-05,
+      "loss": 4.4645,
+      "step": 179200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.70575479096388e-05,
+      "loss": 4.4411,
+      "step": 179712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704917834093201e-05,
+      "loss": 4.4666,
+      "step": 180224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704079239342149e-05,
+      "loss": 4.4423,
+      "step": 180736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.703240644591097e-05,
+      "loss": 4.4411,
+      "step": 181248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.702402049840045e-05,
+      "loss": 4.4417,
+      "step": 181760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7015634550889924e-05,
+      "loss": 4.4398,
+      "step": 182272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7007248603379404e-05,
+      "loss": 4.4483,
+      "step": 182784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.699887903467262e-05,
+      "loss": 4.4412,
+      "step": 183296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.69904930871621e-05,
+      "loss": 4.4304,
+      "step": 183808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6982107139651573e-05,
+      "loss": 4.447,
+      "step": 184320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.697372119214106e-05,
+      "loss": 4.4332,
+      "step": 184832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4379,
+      "step": 185344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6956965675923756e-05,
+      "loss": 4.4335,
+      "step": 185856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694857972841323e-05,
+      "loss": 4.441,
+      "step": 186368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694019378090271e-05,
+      "loss": 4.4202,
+      "step": 186880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.693180783339219e-05,
+      "loss": 4.4471,
+      "step": 187392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4312,
+      "step": 187904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.434,
+      "step": 188416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.4356,
+      "step": 188928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4264,
+      "step": 189440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.688989447464332e-05,
+      "loss": 4.4449,
+      "step": 189952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4484,
+      "step": 190464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.448,
+      "step": 190976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4297,
+      "step": 191488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6856350684601245e-05,
+      "loss": 4.4215,
+      "step": 192000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6847964737090725e-05,
+      "loss": 4.4133,
+      "step": 192512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6839578789580204e-05,
+      "loss": 4.4327,
+      "step": 193024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6831192842069684e-05,
+      "loss": 4.4391,
+      "step": 193536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.682283965216662e-05,
+      "loss": 4.4353,
+      "step": 194048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.4203,
+      "step": 194560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.680608413594931e-05,
+      "loss": 4.412,
+      "step": 195072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.679769818843879e-05,
+      "loss": 4.426,
+      "step": 195584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678931224092827e-05,
+      "loss": 4.414,
+      "step": 196096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678092629341775e-05,
+      "loss": 4.4176,
+      "step": 196608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.677254034590723e-05,
+      "loss": 4.4267,
+      "step": 197120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.676415439839671e-05,
+      "loss": 4.4254,
+      "step": 197632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.675576845088619e-05,
+      "loss": 4.4204,
+      "step": 198144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.4207,
+      "step": 198656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.4087,
+      "step": 199168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4233,
+      "step": 199680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.672224103964785e-05,
+      "loss": 4.417,
+      "step": 200192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.671385509213733e-05,
+      "loss": 4.4033,
+      "step": 200704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.670546914462681e-05,
+      "loss": 4.4053,
+      "step": 201216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.669708319711629e-05,
+      "loss": 4.4221,
+      "step": 201728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668869724960576e-05,
+      "loss": 4.4248,
+      "step": 202240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668031130209524e-05,
+      "loss": 4.4103,
+      "step": 202752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6671941733388456e-05,
+      "loss": 4.411,
+      "step": 203264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6663555785877936e-05,
+      "loss": 4.4166,
+      "step": 203776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6655169838367416e-05,
+      "loss": 4.399,
+      "step": 204288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6646783890856896e-05,
+      "loss": 4.4151,
+      "step": 204800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.4184,
+      "step": 205312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.663002837463959e-05,
+      "loss": 4.4058,
+      "step": 205824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.41,
+      "step": 206336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.3864,
+      "step": 206848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.388,
+      "step": 207360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.3993,
+      "step": 207872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6588098637086985e-05,
+      "loss": 4.4024,
+      "step": 208384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6579712689576465e-05,
+      "loss": 4.4086,
+      "step": 208896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6571326742065945e-05,
+      "loss": 4.3928,
+      "step": 209408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6562957173359154e-05,
+      "loss": 4.3902,
+      "step": 209920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6554571225848634e-05,
+      "loss": 4.3961,
+      "step": 210432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6546185278338114e-05,
+      "loss": 4.4067,
+      "step": 210944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.653781570963133e-05,
+      "loss": 4.3986,
+      "step": 211456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.4015,
+      "step": 211968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652104381461029e-05,
+      "loss": 4.3914,
+      "step": 212480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.651265786709977e-05,
+      "loss": 4.3921,
+      "step": 212992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.650428829839298e-05,
+      "loss": 4.4027,
+      "step": 213504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.649591872968619e-05,
+      "loss": 4.3909,
+      "step": 214016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.648753278217567e-05,
+      "loss": 4.3918,
+      "step": 214528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647914683466515e-05,
+      "loss": 4.3891,
+      "step": 215040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647076088715463e-05,
+      "loss": 4.3801,
+      "step": 215552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.3882,
+      "step": 216064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.645400537093732e-05,
+      "loss": 4.3898,
+      "step": 216576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.64456194234268e-05,
+      "loss": 4.3981,
+      "step": 217088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.643723347591628e-05,
+      "loss": 4.3778,
+      "step": 217600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.642884752840576e-05,
+      "loss": 4.3826,
+      "step": 218112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.3998,
+      "step": 218624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.641207563338472e-05,
+      "loss": 4.3816,
+      "step": 219136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.64036896858742e-05,
+      "loss": 4.3942,
+      "step": 219648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.639530373836368e-05,
+      "loss": 4.3838,
+      "step": 220160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.638691779085316e-05,
+      "loss": 4.4016,
+      "step": 220672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637853184334264e-05,
+      "loss": 4.385,
+      "step": 221184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637014589583212e-05,
+      "loss": 4.3768,
+      "step": 221696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.3761,
+      "step": 222208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.635339037961481e-05,
+      "loss": 4.372,
+      "step": 222720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.634502081090802e-05,
+      "loss": 4.3844,
+      "step": 223232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.633663486339751e-05,
+      "loss": 4.3925,
+      "step": 223744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.632826529469072e-05,
+      "loss": 4.3873,
+      "step": 224256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.63198793471802e-05,
+      "loss": 4.3823,
+      "step": 224768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.631149339966968e-05,
+      "loss": 4.3757,
+      "step": 225280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.630310745215916e-05,
+      "loss": 4.3729,
+      "step": 225792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.629472150464864e-05,
+      "loss": 4.3689,
+      "step": 226304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6286351935941846e-05,
+      "loss": 4.3798,
+      "step": 226816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6277965988431326e-05,
+      "loss": 4.3884,
+      "step": 227328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6269580040920806e-05,
+      "loss": 4.3785,
+      "step": 227840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6261194093410286e-05,
+      "loss": 4.3712,
+      "step": 228352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6252808145899766e-05,
+      "loss": 4.3684,
+      "step": 228864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.344839096069336,
+      "eval_runtime": 537.4924,
+      "eval_samples_per_second": 709.947,
+      "eval_steps_per_second": 22.186,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6244422198389246e-05,
+      "loss": 4.3798,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6236036250878726e-05,
+      "loss": 4.3813,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6227650303368205e-05,
+      "loss": 4.3664,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6219264355857685e-05,
+      "loss": 4.3646,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6210878408347165e-05,
+      "loss": 4.3674,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.3491,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6194106513326125e-05,
+      "loss": 4.3596,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6185720565815605e-05,
+      "loss": 4.3705,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6177334618305085e-05,
+      "loss": 4.3573,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6168948670794565e-05,
+      "loss": 4.367,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160562723284045e-05,
+      "loss": 4.373,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6152176775773525e-05,
+      "loss": 4.3697,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6143790828263005e-05,
+      "loss": 4.3612,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6135404880752485e-05,
+      "loss": 4.3652,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612701893324196e-05,
+      "loss": 4.3584,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118649364535174e-05,
+      "loss": 4.3573,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110263417024654e-05,
+      "loss": 4.3525,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610187746951414e-05,
+      "loss": 4.3601,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093491522003614e-05,
+      "loss": 4.3625,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085105574493094e-05,
+      "loss": 4.364,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6076719626982574e-05,
+      "loss": 4.3554,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6068333679472054e-05,
+      "loss": 4.3656,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6059947731961534e-05,
+      "loss": 4.3554,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605157816325474e-05,
+      "loss": 4.3565,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604319221574422e-05,
+      "loss": 4.3574,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60348062682337e-05,
+      "loss": 4.3529,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602642032072318e-05,
+      "loss": 4.3602,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601805075201639e-05,
+      "loss": 4.3458,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600966480450588e-05,
+      "loss": 4.3594,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600127885699536e-05,
+      "loss": 4.3451,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599289290948484e-05,
+      "loss": 4.3456,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598450696197432e-05,
+      "loss": 4.3495,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3566,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.3441,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595936549824649e-05,
+      "loss": 4.3457,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595097955073597e-05,
+      "loss": 4.3531,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594259360322545e-05,
+      "loss": 4.3458,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593420765571493e-05,
+      "loss": 4.3637,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.592582170820441e-05,
+      "loss": 4.3466,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591743576069389e-05,
+      "loss": 4.3324,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5909066191987096e-05,
+      "loss": 4.3355,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5900680244476576e-05,
+      "loss": 4.3318,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589229429696606e-05,
+      "loss": 4.3529,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588390834945554e-05,
+      "loss": 4.3354,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587552240194502e-05,
+      "loss": 4.3452,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5867136454434496e-05,
+      "loss": 4.3419,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5858750506923976e-05,
+      "loss": 4.3419,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5850364559413456e-05,
+      "loss": 4.349,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584199499070667e-05,
+      "loss": 4.3258,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5833609043196145e-05,
+      "loss": 4.3256,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5825223095685625e-05,
+      "loss": 4.3303,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5816837148175105e-05,
+      "loss": 4.344,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580846757946832e-05,
+      "loss": 4.3225,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58000816319578e-05,
+      "loss": 4.3447,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579169568444728e-05,
+      "loss": 4.324,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.578330973693676e-05,
+      "loss": 4.3197,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577494016822997e-05,
+      "loss": 4.3269,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5766570599523186e-05,
+      "loss": 4.318,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5758184652012666e-05,
+      "loss": 4.3343,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5749798704502146e-05,
+      "loss": 4.3261,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574141275699162e-05,
+      "loss": 4.3141,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57330268094811e-05,
+      "loss": 4.3297,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724657240774315e-05,
+      "loss": 4.3182,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716271293263794e-05,
+      "loss": 4.325,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.570788534575327e-05,
+      "loss": 4.3155,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699499398242754e-05,
+      "loss": 4.3301,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691129829535963e-05,
+      "loss": 4.3077,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5682743882025443e-05,
+      "loss": 4.3303,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674357934514923e-05,
+      "loss": 4.3144,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56659719870044e-05,
+      "loss": 4.318,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565758603949388e-05,
+      "loss": 4.3263,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564920009198336e-05,
+      "loss": 4.32,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564081414447284e-05,
+      "loss": 4.3275,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563244457576605e-05,
+      "loss": 4.3391,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562405862825553e-05,
+      "loss": 4.333,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561568905954874e-05,
+      "loss": 4.317,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560730311203822e-05,
+      "loss": 4.3129,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559891716452771e-05,
+      "loss": 4.3052,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559053121701719e-05,
+      "loss": 4.3225,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55821616483104e-05,
+      "loss": 4.3287,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.3275,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3099,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.3028,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554861785826832e-05,
+      "loss": 4.3214,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55402319107578e-05,
+      "loss": 4.3,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.3123,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.3158,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.3177,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5506704499519446e-05,
+      "loss": 4.3154,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5498318552008926e-05,
+      "loss": 4.3188,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5489932604498406e-05,
+      "loss": 4.2997,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5481546656987886e-05,
+      "loss": 4.3132,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547316070947737e-05,
+      "loss": 4.31,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546479114077058e-05,
+      "loss": 4.3006,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.2985,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3199,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543963329823902e-05,
+      "loss": 4.3127,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54312473507285e-05,
+      "loss": 4.3112,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542286140321798e-05,
+      "loss": 4.3067,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5414475455707455e-05,
+      "loss": 4.314,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5406089508196934e-05,
+      "loss": 4.2972,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5397703560686414e-05,
+      "loss": 4.3102,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389317613175894e-05,
+      "loss": 4.3152,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538094804446911e-05,
+      "loss": 4.3001,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.537256209695859e-05,
+      "loss": 4.3133,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536417614944807e-05,
+      "loss": 4.2813,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535579020193755e-05,
+      "loss": 4.2874,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534740425442703e-05,
+      "loss": 4.2974,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533901830691651e-05,
+      "loss": 4.3006,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533064873820972e-05,
+      "loss": 4.3091,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53222627906992e-05,
+      "loss": 4.291,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531387684318868e-05,
+      "loss": 4.289,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.530549089567816e-05,
+      "loss": 4.2967,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529710494816764e-05,
+      "loss": 4.3064,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528871900065712e-05,
+      "loss": 4.3033,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52803330531466e-05,
+      "loss": 4.3026,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527194710563608e-05,
+      "loss": 4.2904,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5263577536929295e-05,
+      "loss": 4.2923,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255191589418775e-05,
+      "loss": 4.3057,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5246822020711984e-05,
+      "loss": 4.2923,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523845245200519e-05,
+      "loss": 4.2893,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523006650449467e-05,
+      "loss": 4.2938,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.2803,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521329460947363e-05,
+      "loss": 4.2948,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520490866196311e-05,
+      "loss": 4.2933,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519652271445259e-05,
+      "loss": 4.302,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.2822,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517976719823528e-05,
+      "loss": 4.2846,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517138125072476e-05,
+      "loss": 4.3053,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516299530321425e-05,
+      "loss": 4.2863,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515460935570373e-05,
+      "loss": 4.2966,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514622340819321e-05,
+      "loss": 4.2924,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513783746068269e-05,
+      "loss": 4.3035,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512945151317217e-05,
+      "loss": 4.2931,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512106556566164e-05,
+      "loss": 4.2853,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.511269599695486e-05,
+      "loss": 4.2794,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510431004944434e-05,
+      "loss": 4.2778,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509592410193382e-05,
+      "loss": 4.2879,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508753815442329e-05,
+      "loss": 4.2988,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507915220691277e-05,
+      "loss": 4.2929,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507076625940225e-05,
+      "loss": 4.2924,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506238031189173e-05,
+      "loss": 4.2803,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505399436438121e-05,
+      "loss": 4.2817,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5045624795674426e-05,
+      "loss": 4.2741,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503725522696764e-05,
+      "loss": 4.2898,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5028869279457115e-05,
+      "loss": 4.291,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5020483331946595e-05,
+      "loss": 4.2943,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5012097384436075e-05,
+      "loss": 4.2775,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500372781572929e-05,
+      "loss": 4.2765,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.261576175689697,
+      "eval_runtime": 548.1325,
+      "eval_samples_per_second": 696.166,
+      "eval_steps_per_second": 21.756,
+      "step": 305276
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4995341868218764e-05,
+      "loss": 4.2887,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4986955920708244e-05,
+      "loss": 4.2862,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978569973197724e-05,
+      "loss": 4.2749,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4970184025687204e-05,
+      "loss": 4.2735,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4961798078176684e-05,
+      "loss": 4.2758,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4953412130666164e-05,
+      "loss": 4.2645,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494504256195938e-05,
+      "loss": 4.2729,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.493665661444886e-05,
+      "loss": 4.2754,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492827066693834e-05,
+      "loss": 4.2722,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491990109823155e-05,
+      "loss": 4.2798,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491151515072103e-05,
+      "loss": 4.2827,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490312920321051e-05,
+      "loss": 4.2753,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489474325569999e-05,
+      "loss": 4.2753,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.488635730818947e-05,
+      "loss": 4.2784,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487797136067895e-05,
+      "loss": 4.2731,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486958541316843e-05,
+      "loss": 4.2678,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486119946565791e-05,
+      "loss": 4.2625,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485282989695112e-05,
+      "loss": 4.278,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4844443949440604e-05,
+      "loss": 4.2718,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4836058001930084e-05,
+      "loss": 4.2801,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4827672054419564e-05,
+      "loss": 4.2673,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4819286106909044e-05,
+      "loss": 4.2777,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4810900159398524e-05,
+      "loss": 4.2716,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4802514211888004e-05,
+      "loss": 4.271,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.479412826437748e-05,
+      "loss": 4.2702,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.478574231686696e-05,
+      "loss": 4.2673,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.477737274816017e-05,
+      "loss": 4.2749,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476898680064965e-05,
+      "loss": 4.259,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4760600853139126e-05,
+      "loss": 4.2733,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4752214905628606e-05,
+      "loss": 4.2615,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474384533692182e-05,
+      "loss": 4.2575,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.47354593894113e-05,
+      "loss": 4.2667,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472707344190078e-05,
+      "loss": 4.2724,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471868749439026e-05,
+      "loss": 4.2646,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471030154687974e-05,
+      "loss": 4.2599,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470193197817295e-05,
+      "loss": 4.2657,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469354603066243e-05,
+      "loss": 4.2673,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468516008315191e-05,
+      "loss": 4.2797,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467677413564139e-05,
+      "loss": 4.2629,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466838818813087e-05,
+      "loss": 4.251,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466000224062035e-05,
+      "loss": 4.2514,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.465161629310983e-05,
+      "loss": 4.2541,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.464323034559931e-05,
+      "loss": 4.2674,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.463484439808879e-05,
+      "loss": 4.2557,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4626474829382006e-05,
+      "loss": 4.2656,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618088881871486e-05,
+      "loss": 4.2613,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609702934360966e-05,
+      "loss": 4.2576,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4601316986850446e-05,
+      "loss": 4.2692,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4592947418143655e-05,
+      "loss": 4.2478,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4584561470633135e-05,
+      "loss": 4.2409,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576175523122615e-05,
+      "loss": 4.2522,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4567789575612095e-05,
+      "loss": 4.2589,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4559420006905304e-05,
+      "loss": 4.2418,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4551034059394784e-05,
+      "loss": 4.2652,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4542648111884264e-05,
+      "loss": 4.245,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4534262164373744e-05,
+      "loss": 4.241,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452589259566696e-05,
+      "loss": 4.2469,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451750664815644e-05,
+      "loss": 4.2409,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450912070064592e-05,
+      "loss": 4.2578,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45007347531354e-05,
+      "loss": 4.2507,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.449238156323234e-05,
+      "loss": 4.2351,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.448399561572182e-05,
+      "loss": 4.2483,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44756096682113e-05,
+      "loss": 4.2404,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446722372070078e-05,
+      "loss": 4.2528,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445883777319026e-05,
+      "loss": 4.2379,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445045182567974e-05,
+      "loss": 4.2505,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.444206587816922e-05,
+      "loss": 4.2307,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44336799306587e-05,
+      "loss": 4.2509,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4425310361951914e-05,
+      "loss": 4.2411,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.441694079324512e-05,
+      "loss": 4.2392,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44085548457346e-05,
+      "loss": 4.2502,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440016889822408e-05,
+      "loss": 4.2407,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.439178295071356e-05,
+      "loss": 4.2522,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.438339700320304e-05,
+      "loss": 4.2601,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.437501105569252e-05,
+      "loss": 4.26,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.436664148698573e-05,
+      "loss": 4.2422,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435825553947521e-05,
+      "loss": 4.2383,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434986959196469e-05,
+      "loss": 4.2298,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434148364445417e-05,
+      "loss": 4.2448,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433309769694365e-05,
+      "loss": 4.2565,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432471174943313e-05,
+      "loss": 4.2488,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431632580192261e-05,
+      "loss": 4.2378,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.430793985441209e-05,
+      "loss": 4.2251,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429957028570531e-05,
+      "loss": 4.2514,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429118433819479e-05,
+      "loss": 4.2265,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428279839068426e-05,
+      "loss": 4.2358,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4274428821977476e-05,
+      "loss": 4.2398,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4266042874466956e-05,
+      "loss": 4.2416,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.425765692695643e-05,
+      "loss": 4.2439,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424927097944591e-05,
+      "loss": 4.2443,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424088503193539e-05,
+      "loss": 4.2283,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423249908442487e-05,
+      "loss": 4.2405,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422411313691435e-05,
+      "loss": 4.2365,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4215743568207565e-05,
+      "loss": 4.2279,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4207357620697045e-05,
+      "loss": 4.2254,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4198971673186525e-05,
+      "loss": 4.2489,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4190585725676005e-05,
+      "loss": 4.2411,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4182199778165485e-05,
+      "loss": 4.241,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4173813830654965e-05,
+      "loss": 4.2348,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4165444261948174e-05,
+      "loss": 4.2398,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4157058314437654e-05,
+      "loss": 4.2258,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4148672366927134e-05,
+      "loss": 4.2401,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4140286419416614e-05,
+      "loss": 4.2412,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.413191685070982e-05,
+      "loss": 4.2304,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.41235309031993e-05,
+      "loss": 4.2438,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411514495568879e-05,
+      "loss": 4.212,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.410675900817827e-05,
+      "loss": 4.2184,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409837306066775e-05,
+      "loss": 4.225,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408998711315723e-05,
+      "loss": 4.2305,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408160116564671e-05,
+      "loss": 4.2389,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407323159693992e-05,
+      "loss": 4.2251,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406486202823313e-05,
+      "loss": 4.2168,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405647608072261e-05,
+      "loss": 4.2236,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404809013321209e-05,
+      "loss": 4.236,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403970418570157e-05,
+      "loss": 4.2338,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403131823819105e-05,
+      "loss": 4.235,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.402293229068053e-05,
+      "loss": 4.2216,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401454634317001e-05,
+      "loss": 4.2225,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400616039565949e-05,
+      "loss": 4.2369,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.399777444814897e-05,
+      "loss": 4.2259,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398938850063845e-05,
+      "loss": 4.2208,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398100255312793e-05,
+      "loss": 4.2249,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397261660561741e-05,
+      "loss": 4.2137,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396426341571435e-05,
+      "loss": 4.2273,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.395587746820383e-05,
+      "loss": 4.2265,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394749152069331e-05,
+      "loss": 4.2335,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393910557318279e-05,
+      "loss": 4.214,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3930719625672265e-05,
+      "loss": 4.2178,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3922333678161745e-05,
+      "loss": 4.2334,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3913947730651225e-05,
+      "loss": 4.2221,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3905561783140705e-05,
+      "loss": 4.2281,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389719221443392e-05,
+      "loss": 4.2279,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38888062669234e-05,
+      "loss": 4.2333,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388043669821662e-05,
+      "loss": 4.2303,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2165,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2147,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.2168,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3846909286978266e-05,
+      "loss": 4.2222,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383852333946774e-05,
+      "loss": 4.2271,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383013739195722e-05,
+      "loss": 4.2284,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38217514444467e-05,
+      "loss": 4.2264,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381336549693618e-05,
+      "loss": 4.2178,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.2179,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379662635952261e-05,
+      "loss": 4.2078,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378824041201209e-05,
+      "loss": 4.2255,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.2271,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.229,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3763082569480524e-05,
+      "loss": 4.212,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3754696621970004e-05,
+      "loss": 4.2121,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.205011367797852,
+      "eval_runtime": 548.3436,
+      "eval_samples_per_second": 695.898,
+      "eval_steps_per_second": 21.747,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3746310674459483e-05,
+      "loss": 4.2355,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3737924726948963e-05,
+      "loss": 4.2182,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3729538779438443e-05,
+      "loss": 4.2134,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372115283192792e-05,
+      "loss": 4.2076,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37127668844174e-05,
+      "loss": 4.2117,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370438093690688e-05,
+      "loss": 4.2014,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369599498939636e-05,
+      "loss": 4.21,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368760904188584e-05,
+      "loss": 4.2063,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367922309437533e-05,
+      "loss": 4.2098,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367083714686481e-05,
+      "loss": 4.2202,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366245119935428e-05,
+      "loss": 4.2143,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365406525184376e-05,
+      "loss": 4.2109,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364567930433324e-05,
+      "loss": 4.2159,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363729335682272e-05,
+      "loss": 4.2148,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36289074093122e-05,
+      "loss": 4.2091,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362052146180168e-05,
+      "loss": 4.2028,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361213551429116e-05,
+      "loss": 4.2008,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360376594558437e-05,
+      "loss": 4.2095,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359537999807385e-05,
+      "loss": 4.212,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358699405056333e-05,
+      "loss": 4.2196,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357860810305281e-05,
+      "loss": 4.2069,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35702221555423e-05,
+      "loss": 4.2155,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356183620803178e-05,
+      "loss": 4.2122,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355346663932499e-05,
+      "loss": 4.2088,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354508069181447e-05,
+      "loss": 4.2073,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353669474430395e-05,
+      "loss": 4.2054,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528325175597157e-05,
+      "loss": 4.215,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3519939228086636e-05,
+      "loss": 4.1967,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511553280576116e-05,
+      "loss": 4.217,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503167333065596e-05,
+      "loss": 4.2006,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3494781385555076e-05,
+      "loss": 4.1952,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486395438044556e-05,
+      "loss": 4.2085,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3478009490534036e-05,
+      "loss": 4.2083,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3469623543023516e-05,
+      "loss": 4.2102,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461237595512996e-05,
+      "loss": 4.201,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3452851648002476e-05,
+      "loss": 4.201,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344448207929569e-05,
+      "loss": 4.2074,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436096131785165e-05,
+      "loss": 4.2207,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3427710184274645e-05,
+      "loss": 4.2051,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419324236764125e-05,
+      "loss": 4.1931,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3410938289253605e-05,
+      "loss": 4.1935,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402568720546814e-05,
+      "loss": 4.1921,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394182773036294e-05,
+      "loss": 4.2046,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385796825525774e-05,
+      "loss": 4.1996,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377410878015254e-05,
+      "loss": 4.205,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336904130930847e-05,
+      "loss": 4.2053,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336065536179795e-05,
+      "loss": 4.1995,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335226941428743e-05,
+      "loss": 4.2102,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334388346677691e-05,
+      "loss": 4.189,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333549751926639e-05,
+      "loss": 4.1819,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332711157175587e-05,
+      "loss": 4.1973,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331874200304908e-05,
+      "loss": 4.1936,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331035605553856e-05,
+      "loss": 4.1872,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330197010802804e-05,
+      "loss": 4.2049,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329360053932125e-05,
+      "loss": 4.1855,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328521459181073e-05,
+      "loss": 4.1852,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327682864430021e-05,
+      "loss": 4.1878,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326844269678969e-05,
+      "loss": 4.1842,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326005674927917e-05,
+      "loss": 4.1991,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3251670801768654e-05,
+      "loss": 4.1895,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243284854258134e-05,
+      "loss": 4.1858,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3234898906747614e-05,
+      "loss": 4.1862,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3226512959237094e-05,
+      "loss": 4.1777,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3218127011726574e-05,
+      "loss": 4.1989,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320975744301978e-05,
+      "loss": 4.1812,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320137149550926e-05,
+      "loss": 4.2003,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319298554799874e-05,
+      "loss": 4.1708,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318459960048822e-05,
+      "loss": 4.1919,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31762136529777e-05,
+      "loss": 4.185,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3167827705467176e-05,
+      "loss": 4.1819,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3159441757956656e-05,
+      "loss": 4.1938,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151055810446136e-05,
+      "loss": 4.1863,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314270262054309e-05,
+      "loss": 4.1927,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313431667303257e-05,
+      "loss": 4.2025,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312593072552205e-05,
+      "loss": 4.2027,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311754477801153e-05,
+      "loss": 4.1915,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109158830501e-05,
+      "loss": 4.1826,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310077288299048e-05,
+      "loss": 4.1755,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30924033142837e-05,
+      "loss": 4.1885,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308401736677318e-05,
+      "loss": 4.1981,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307563141926265e-05,
+      "loss": 4.1929,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306724547175213e-05,
+      "loss": 4.184,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305885952424161e-05,
+      "loss": 4.1699,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305047357673109e-05,
+      "loss": 4.1975,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042087629220577e-05,
+      "loss": 4.1677,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3033701681710057e-05,
+      "loss": 4.1839,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025315734199536e-05,
+      "loss": 4.1848,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3016929786689016e-05,
+      "loss": 4.19,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008560217982226e-05,
+      "loss": 4.1872,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000174270471705e-05,
+      "loss": 4.1957,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2991788322961185e-05,
+      "loss": 4.1711,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983402375450665e-05,
+      "loss": 4.1816,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975016427940145e-05,
+      "loss": 4.1854,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2966630480429625e-05,
+      "loss": 4.1757,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.1724,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949874964212314e-05,
+      "loss": 4.1918,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2941489016701794e-05,
+      "loss": 4.1854,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293311944799501e-05,
+      "loss": 4.1936,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292473350048449e-05,
+      "loss": 4.1788,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291634755297397e-05,
+      "loss": 4.1859,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290796160546345e-05,
+      "loss": 4.17,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289957565795293e-05,
+      "loss": 4.1876,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289118971044241e-05,
+      "loss": 4.1906,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288280376293189e-05,
+      "loss": 4.1754,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287441781542136e-05,
+      "loss": 4.1902,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286604824671458e-05,
+      "loss": 4.1572,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285766229920406e-05,
+      "loss": 4.1699,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284927635169354e-05,
+      "loss": 4.1681,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284089040418301e-05,
+      "loss": 4.1786,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283250445667249e-05,
+      "loss": 4.1869,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282413488796571e-05,
+      "loss": 4.178,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281574894045519e-05,
+      "loss": 4.1569,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807379371748404e-05,
+      "loss": 4.1776,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2798993424237884e-05,
+      "loss": 4.1829,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2790607476727364e-05,
+      "loss": 4.1833,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278222152921684e-05,
+      "loss": 4.18,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277383558170632e-05,
+      "loss": 4.1705,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27654496341958e-05,
+      "loss": 4.1686,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275706368668528e-05,
+      "loss": 4.1856,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274867773917476e-05,
+      "loss": 4.1723,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2740291791664237e-05,
+      "loss": 4.1723,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2731922222957446e-05,
+      "loss": 4.1686,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272353627544693e-05,
+      "loss": 4.1641,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271515032793641e-05,
+      "loss": 4.1738,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270676438042589e-05,
+      "loss": 4.1761,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269837843291537e-05,
+      "loss": 4.1821,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268999248540485e-05,
+      "loss": 4.1661,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268160653789433e-05,
+      "loss": 4.1672,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267322059038381e-05,
+      "loss": 4.1842,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266483464287329e-05,
+      "loss": 4.1685,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26564650741665e-05,
+      "loss": 4.175,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264807912665598e-05,
+      "loss": 4.1787,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263970955794919e-05,
+      "loss": 4.1855,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263132361043867e-05,
+      "loss": 4.1806,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262293766292815e-05,
+      "loss": 4.1674,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261455171541763e-05,
+      "loss": 4.1586,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260616576790712e-05,
+      "loss": 4.167,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25977798203966e-05,
+      "loss": 4.1715,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258939387288608e-05,
+      "loss": 4.1819,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2581024304179286e-05,
+      "loss": 4.176,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2572654735472495e-05,
+      "loss": 4.1723,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564268787961975e-05,
+      "loss": 4.1723,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2555882840451455e-05,
+      "loss": 4.1711,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2547496892940935e-05,
+      "loss": 4.1582,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2539110945430415e-05,
+      "loss": 4.1689,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2530724997919895e-05,
+      "loss": 4.178,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2522339050409375e-05,
+      "loss": 4.1812,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2513953102898855e-05,
+      "loss": 4.1614,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2505567155388335e-05,
+      "loss": 4.1608,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.163527965545654,
+      "eval_runtime": 549.5824,
+      "eval_samples_per_second": 694.329,
+      "eval_steps_per_second": 21.698,
+      "step": 457914
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249719758668155e-05,
+      "loss": 4.1819,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2488811639171024e-05,
+      "loss": 4.1695,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2480425691660504e-05,
+      "loss": 4.162,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2472039744149984e-05,
+      "loss": 4.1619,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2463653796639464e-05,
+      "loss": 4.159,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2455267849128943e-05,
+      "loss": 4.1541,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244689828042215e-05,
+      "loss": 4.1578,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243851233291163e-05,
+      "loss": 4.1597,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243012638540111e-05,
+      "loss": 4.164,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242174043789059e-05,
+      "loss": 4.1682,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241337086918381e-05,
+      "loss": 4.1687,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240498492167329e-05,
+      "loss": 4.1577,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23966153529665e-05,
+      "loss": 4.1733,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238822940545598e-05,
+      "loss": 4.1634,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237984345794546e-05,
+      "loss": 4.1653,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237145751043494e-05,
+      "loss": 4.1517,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236307156292442e-05,
+      "loss": 4.1523,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23546856154139e-05,
+      "loss": 4.1658,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234629966790338e-05,
+      "loss": 4.1626,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233791372039286e-05,
+      "loss": 4.1732,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232952777288234e-05,
+      "loss": 4.1593,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232114182537182e-05,
+      "loss": 4.1684,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23127558778613e-05,
+      "loss": 4.1655,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230436993035078e-05,
+      "loss": 4.1628,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229598398284026e-05,
+      "loss": 4.1587,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228761441413347e-05,
+      "loss": 4.1571,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227922846662295e-05,
+      "loss": 4.1669,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227084251911243e-05,
+      "loss": 4.1502,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226245657160191e-05,
+      "loss": 4.1683,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2254070624091386e-05,
+      "loss": 4.1555,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2245684676580866e-05,
+      "loss": 4.1473,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223731510787408e-05,
+      "loss": 4.1605,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222892916036356e-05,
+      "loss": 4.163,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2220543212853035e-05,
+      "loss": 4.1681,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2212157265342515e-05,
+      "loss": 4.1521,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2203771317831995e-05,
+      "loss": 4.1577,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219540174912521e-05,
+      "loss": 4.1612,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218701580161469e-05,
+      "loss": 4.1703,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217862985410417e-05,
+      "loss": 4.1614,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217024390659365e-05,
+      "loss": 4.1453,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216185795908313e-05,
+      "loss": 4.1463,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215347201157261e-05,
+      "loss": 4.1447,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214510244286582e-05,
+      "loss": 4.1643,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21367164953553e-05,
+      "loss": 4.1549,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212833054784478e-05,
+      "loss": 4.1537,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211994460033426e-05,
+      "loss": 4.164,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211155865282374e-05,
+      "loss": 4.1529,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210317270531322e-05,
+      "loss": 4.1634,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20947867578027e-05,
+      "loss": 4.1454,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208641718909591e-05,
+      "loss": 4.1394,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2078031241585395e-05,
+      "loss": 4.1504,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2069645294074875e-05,
+      "loss": 4.1473,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061275725368084e-05,
+      "loss": 4.1476,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2052889777857564e-05,
+      "loss": 4.1543,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2044503830347044e-05,
+      "loss": 4.1448,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036117882836524e-05,
+      "loss": 4.14,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2027731935326004e-05,
+      "loss": 4.1409,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019345987815484e-05,
+      "loss": 4.1417,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2010960040304964e-05,
+      "loss": 4.1516,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2002574092794444e-05,
+      "loss": 4.1448,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199420452408765e-05,
+      "loss": 4.1398,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198581857657713e-05,
+      "loss": 4.1433,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197743262906661e-05,
+      "loss": 4.1315,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196904668155609e-05,
+      "loss": 4.1579,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196067711284931e-05,
+      "loss": 4.1397,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195230754414252e-05,
+      "loss": 4.1491,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1943921596632e-05,
+      "loss": 4.1355,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193553564912148e-05,
+      "loss": 4.1412,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192714970161096e-05,
+      "loss": 4.1443,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191876375410044e-05,
+      "loss": 4.1387,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191037780658992e-05,
+      "loss": 4.1506,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19019918590794e-05,
+      "loss": 4.142,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1893622290372606e-05,
+      "loss": 4.1493,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1885236342862086e-05,
+      "loss": 4.1582,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1876850395351566e-05,
+      "loss": 4.1548,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1868464447841046e-05,
+      "loss": 4.1498,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1860078500330526e-05,
+      "loss": 4.1418,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1851692552820006e-05,
+      "loss": 4.1333,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184332298411322e-05,
+      "loss": 4.1444,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1834937036602695e-05,
+      "loss": 4.1538,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1826551089092175e-05,
+      "loss": 4.1503,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1818165141581655e-05,
+      "loss": 4.1422,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1809779194071135e-05,
+      "loss": 4.1275,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1801393246560615e-05,
+      "loss": 4.1555,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1793007299050095e-05,
+      "loss": 4.1229,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1784621351539575e-05,
+      "loss": 4.1419,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776235404029055e-05,
+      "loss": 4.1438,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1767865835322264e-05,
+      "loss": 4.1425,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.175947988781175e-05,
+      "loss": 4.1481,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175109394030123e-05,
+      "loss": 4.1519,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174272437159444e-05,
+      "loss": 4.1258,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173433842408392e-05,
+      "loss": 4.1416,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17259524765734e-05,
+      "loss": 4.1432,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171756652906288e-05,
+      "loss": 4.1336,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170918058155236e-05,
+      "loss": 4.128,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170079463404184e-05,
+      "loss": 4.1495,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169240868653132e-05,
+      "loss": 4.1441,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16840227390208e-05,
+      "loss": 4.1529,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167565317031401e-05,
+      "loss": 4.1354,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166726722280349e-05,
+      "loss": 4.1442,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165888127529297e-05,
+      "loss": 4.1286,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165049532778245e-05,
+      "loss": 4.1445,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1642125759075664e-05,
+      "loss": 4.1473,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1633739811565144e-05,
+      "loss": 4.1409,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1625353864054624e-05,
+      "loss": 4.1454,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1616967916544104e-05,
+      "loss": 4.1189,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160859834783731e-05,
+      "loss": 4.1243,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160021240032679e-05,
+      "loss": 4.1244,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159184283162e-05,
+      "loss": 4.1369,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158345688410948e-05,
+      "loss": 4.1458,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157507093659896e-05,
+      "loss": 4.1421,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156668498908844e-05,
+      "loss": 4.1149,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155831542038166e-05,
+      "loss": 4.1338,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154992947287114e-05,
+      "loss": 4.1425,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154154352536062e-05,
+      "loss": 4.1435,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15331575778501e-05,
+      "loss": 4.1394,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152477163033958e-05,
+      "loss": 4.1303,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151638568282906e-05,
+      "loss": 4.1314,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150799973531853e-05,
+      "loss": 4.1411,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149963016661175e-05,
+      "loss": 4.1358,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149124421910123e-05,
+      "loss": 4.13,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14828582715907e-05,
+      "loss": 4.1278,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147447232408018e-05,
+      "loss": 4.123,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146608637656966e-05,
+      "loss": 4.1362,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145770042905914e-05,
+      "loss": 4.1355,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144931448154863e-05,
+      "loss": 4.1384,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144092853403811e-05,
+      "loss": 4.1291,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1432542586527587e-05,
+      "loss": 4.1251,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1424173017820796e-05,
+      "loss": 4.1438,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1415787070310276e-05,
+      "loss": 4.1247,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1407401122799756e-05,
+      "loss": 4.1384,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1399015175289236e-05,
+      "loss": 4.1429,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1390645606582445e-05,
+      "loss": 4.1412,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382259659071925e-05,
+      "loss": 4.1413,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1373873711561405e-05,
+      "loss": 4.133,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1365487764050885e-05,
+      "loss": 4.1176,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357118195344094e-05,
+      "loss": 4.1243,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134873224783358e-05,
+      "loss": 4.131,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134034630032306e-05,
+      "loss": 4.1449,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133196035281254e-05,
+      "loss": 4.134,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132357440530202e-05,
+      "loss": 4.1379,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131520483659523e-05,
+      "loss": 4.1296,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130681888908471e-05,
+      "loss": 4.1313,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129843294157419e-05,
+      "loss": 4.1177,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12900633728674e-05,
+      "loss": 4.1308,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128167742535688e-05,
+      "loss": 4.1403,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127329147784636e-05,
+      "loss": 4.1422,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126490553033584e-05,
+      "loss": 4.1197,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125651958282532e-05,
+      "loss": 4.1255,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.132448196411133,
+      "eval_runtime": 550.9379,
+      "eval_samples_per_second": 692.621,
+      "eval_steps_per_second": 21.645,
+      "step": 534233
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12481336353148e-05,
+      "loss": 4.1517,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123974768780428e-05,
+      "loss": 4.1312,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123136174029376e-05,
+      "loss": 4.1247,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1222975792783245e-05,
+      "loss": 4.1268,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121458984527272e-05,
+      "loss": 4.1168,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12062038977622e-05,
+      "loss": 4.1214,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119781795025168e-05,
+      "loss": 4.1139,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118943200274116e-05,
+      "loss": 4.125,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118104605523064e-05,
+      "loss": 4.1215,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117266010772012e-05,
+      "loss": 4.1308,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11642741602096e-05,
+      "loss": 4.1291,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115588821269908e-05,
+      "loss": 4.1236,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114750226518856e-05,
+      "loss": 4.1349,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113911631767804e-05,
+      "loss": 4.1217,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113073037016752e-05,
+      "loss": 4.132,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122344422657e-05,
+      "loss": 4.1106,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111397485395021e-05,
+      "loss": 4.1185,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110558890643969e-05,
+      "loss": 4.1267,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109720295892917e-05,
+      "loss": 4.1233,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108881701141865e-05,
+      "loss": 4.1346,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108044744271186e-05,
+      "loss": 4.1228,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107206149520134e-05,
+      "loss": 4.1292,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106367554769082e-05,
+      "loss": 4.1277,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10552896001803e-05,
+      "loss": 4.1224,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1046903652669775e-05,
+      "loss": 4.1261,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1038517705159255e-05,
+      "loss": 4.1154,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030131757648735e-05,
+      "loss": 4.1284,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1021745810138215e-05,
+      "loss": 4.1196,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1013359862627695e-05,
+      "loss": 4.1294,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100497391511718e-05,
+      "loss": 4.1139,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099658796760666e-05,
+      "loss": 4.1127,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098820202009614e-05,
+      "loss": 4.1229,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097981607258562e-05,
+      "loss": 4.1239,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097144650387883e-05,
+      "loss": 4.1282,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096306055636831e-05,
+      "loss": 4.1218,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095467460885779e-05,
+      "loss": 4.1149,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946305040151e-05,
+      "loss": 4.1248,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093791909264048e-05,
+      "loss": 4.1309,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092953314512996e-05,
+      "loss": 4.1237,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092114719761944e-05,
+      "loss": 4.1135,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091276125010892e-05,
+      "loss": 4.1079,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09043753025984e-05,
+      "loss": 4.1096,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089598935508788e-05,
+      "loss": 4.1258,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088760340757736e-05,
+      "loss": 4.1197,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879233838870575e-05,
+      "loss": 4.1145,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0870864270163784e-05,
+      "loss": 4.1309,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0862478322653264e-05,
+      "loss": 4.1144,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0854092375142744e-05,
+      "loss": 4.1277,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0845706427632224e-05,
+      "loss": 4.1101,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0837320480121704e-05,
+      "loss": 4.1024,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0828950911414913e-05,
+      "loss": 4.112,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082056496390439e-05,
+      "loss": 4.1144,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081217901639387e-05,
+      "loss": 4.1064,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080380944768708e-05,
+      "loss": 4.1208,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079542350017657e-05,
+      "loss": 4.1107,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078703755266605e-05,
+      "loss": 4.1007,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077865160515553e-05,
+      "loss": 4.1109,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077026565764501e-05,
+      "loss": 4.1006,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076187971013449e-05,
+      "loss": 4.1156,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075349376262397e-05,
+      "loss": 4.1098,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074510781511344e-05,
+      "loss": 4.1073,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073672186760292e-05,
+      "loss": 4.1063,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07283359200924e-05,
+      "loss": 4.0985,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071994997258188e-05,
+      "loss": 4.1191,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071158040387509e-05,
+      "loss": 4.1025,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070319445636457e-05,
+      "loss": 4.1153,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069480850885405e-05,
+      "loss": 4.101,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068642256134354e-05,
+      "loss": 4.1066,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067803661383302e-05,
+      "loss": 4.1048,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066966704512623e-05,
+      "loss": 4.1063,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066128109761571e-05,
+      "loss": 4.1128,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065289515010519e-05,
+      "loss": 4.1076,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064450920259467e-05,
+      "loss": 4.1123,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636139633887876e-05,
+      "loss": 4.1235,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0627753686377356e-05,
+      "loss": 4.121,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0619367738866836e-05,
+      "loss": 4.1117,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0610981791356316e-05,
+      "loss": 4.1122,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0602595843845796e-05,
+      "loss": 4.0926,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594209896335276e-05,
+      "loss": 4.1084,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0585823948824755e-05,
+      "loss": 4.1218,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0577438001314235e-05,
+      "loss": 4.1131,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056905205380372e-05,
+      "loss": 4.1122,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056068248509693e-05,
+      "loss": 4.0931,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055229653758641e-05,
+      "loss": 4.1154,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054391059007589e-05,
+      "loss": 4.0923,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053552464256537e-05,
+      "loss": 4.1052,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052715507385858e-05,
+      "loss": 4.1037,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051876912634806e-05,
+      "loss": 4.1113,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051038317883754e-05,
+      "loss": 4.1129,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050201361013075e-05,
+      "loss": 4.1138,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049362766262023e-05,
+      "loss": 4.0948,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048524171510971e-05,
+      "loss": 4.1123,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047685576759919e-05,
+      "loss": 4.1043,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0468469820088676e-05,
+      "loss": 4.101,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460083872578156e-05,
+      "loss": 4.0931,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045169792506763e-05,
+      "loss": 4.1138,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044331197755711e-05,
+      "loss": 4.1076,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0434958787654054e-05,
+      "loss": 4.1178,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0426572840143534e-05,
+      "loss": 4.106,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418186892633014e-05,
+      "loss": 4.1095,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0409800945122494e-05,
+      "loss": 4.0925,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0401414997611974e-05,
+      "loss": 4.1135,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039304542890518e-05,
+      "loss": 4.1117,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03846758601984e-05,
+      "loss": 4.1096,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037628991268788e-05,
+      "loss": 4.1088,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036790396517736e-05,
+      "loss": 4.0904,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035951801766684e-05,
+      "loss": 4.0902,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035113207015632e-05,
+      "loss": 4.0933,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03427461226458e-05,
+      "loss": 4.1,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033436017513527e-05,
+      "loss": 4.1133,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032597422762475e-05,
+      "loss": 4.1064,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031758828011423e-05,
+      "loss": 4.0781,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030920233260371e-05,
+      "loss": 4.1049,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030081638509319e-05,
+      "loss": 4.1114,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029243043758267e-05,
+      "loss": 4.1053,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028406086887588e-05,
+      "loss": 4.1072,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027567492136537e-05,
+      "loss": 4.1011,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0267305352658576e-05,
+      "loss": 4.0933,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0258919405148056e-05,
+      "loss": 4.108,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0250533457637536e-05,
+      "loss": 4.1051,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242147510127016e-05,
+      "loss": 4.0976,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0233761562616496e-05,
+      "loss": 4.0914,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0225375615105976e-05,
+      "loss": 4.0949,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0216989667595456e-05,
+      "loss": 4.0999,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0208603720084936e-05,
+      "loss": 4.1009,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0200217772574416e-05,
+      "loss": 4.1096,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0191831825063896e-05,
+      "loss": 4.0953,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0183462256357105e-05,
+      "loss": 4.0917,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175076308846585e-05,
+      "loss": 4.1119,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0166690361336065e-05,
+      "loss": 4.0903,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0158304413825545e-05,
+      "loss": 4.1068,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014995122392249e-05,
+      "loss": 4.11,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014156527641197e-05,
+      "loss": 4.1057,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013317932890145e-05,
+      "loss": 4.11,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012479338139093e-05,
+      "loss": 4.0992,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011640743388041e-05,
+      "loss": 4.0888,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010802148636989e-05,
+      "loss": 4.0923,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009963553885937e-05,
+      "loss": 4.1016,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009124959134885e-05,
+      "loss": 4.1076,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008286364383833e-05,
+      "loss": 4.1043,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007449407513154e-05,
+      "loss": 4.1037,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006610812762102e-05,
+      "loss": 4.0953,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00577221801105e-05,
+      "loss": 4.0989,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0049336232599985e-05,
+      "loss": 4.0853,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0040966663893194e-05,
+      "loss": 4.1008,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0032580716382674e-05,
+      "loss": 4.1068,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0024194768872154e-05,
+      "loss": 4.1109,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0015825200165363e-05,
+      "loss": 4.0858,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0007439252654843e-05,
+      "loss": 4.0959,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.107790470123291,
+      "eval_runtime": 548.772,
+      "eval_samples_per_second": 695.354,
+      "eval_steps_per_second": 21.73,
+      "step": 610552
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.519804656749205e+17,
+  "trial_name": null,
+  "trial_params": null
+}